LLMsForHepth/cs_primary_200k · Datasets at Hugging Face

id stringlengths 9 10	submitter stringlengths 1 64 ⌀	authors stringlengths 4 20.7k	title stringlengths 4 246	comments stringlengths 1 523 ⌀	journal-ref stringlengths 4 404 ⌀	doi stringlengths 11 153 ⌀	report-no stringlengths 2 254 ⌀	categories stringlengths 5 98	license stringclasses 9 values	orig_abstract stringlengths 14 3.35k	versions listlengths 1 60	update_date stringlengths 10 10	authors_parsed sequencelengths 1 1.35k	abstract stringlengths 11 3.34k
1906.10370	Miguel Sepulcre	Rafael Molina-Masegosa, Miguel Sepulcre and Javier Gozalvez	Geo-Based Scheduling for C-V2X Networks	null	IEEE Transactions on Vehicular Technology, Volume 68, Issue 9, pp. 8397 - 8407, September 2019	10.1109/TVT.2019.2924698	null	cs.NI	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Cellular Vehicle-to-Everything (C-V2X) networks can operate without cellular infrastructure support. Vehicles can autonomously select their radio resources using the sensing-based Semi-Persistent Scheduling (SPS) algorithm specified by the Third Generation Partnership Project (3GPP). The sensing nature of the SPS scheme makes C-V2X communications prone to the well-known hidden-terminal problem. To address this problem, this paper proposes a novel geo-based scheduling scheme that allows vehicles to autonomously select their radio resources based on the location and ordering of neighboring vehicles on the road. The proposed scheme results in an implicit resource selection coordination between vehicles (even with those outside the sensing range) that reduces packet collisions. This paper evaluates analytically and through simulations the proposed scheduling scheme. The obtained results demonstrate that it reduces packet collisions and significantly increases the C-V2X performance compared to when using the sensing-based SPS scheme.	[ { "created": "Tue, 25 Jun 2019 08:12:28 GMT", "version": "v1" } ]	2019-09-20	[ [ "Molina-Masegosa", "Rafael", "" ], [ "Sepulcre", "Miguel", "" ], [ "Gozalvez", "Javier", "" ] ]	Cellular Vehicle-to-Everything (C-V2X) networks can operate without cellular infrastructure support. Vehicles can autonomously select their radio resources using the sensing-based Semi-Persistent Scheduling (SPS) algorithm specified by the Third Generation Partnership Project (3GPP). The sensing nature of the SPS scheme makes C-V2X communications prone to the well-known hidden-terminal problem. To address this problem, this paper proposes a novel geo-based scheduling scheme that allows vehicles to autonomously select their radio resources based on the location and ordering of neighboring vehicles on the road. The proposed scheme results in an implicit resource selection coordination between vehicles (even with those outside the sensing range) that reduces packet collisions. This paper evaluates analytically and through simulations the proposed scheduling scheme. The obtained results demonstrate that it reduces packet collisions and significantly increases the C-V2X performance compared to when using the sensing-based SPS scheme.
2312.07340	Yusen Feng	Yusen Feng, Xiyan Xu, Libin Liu	MuscleVAE: Model-Based Controllers of Muscle-Actuated Characters	null	null	null	null	cs.GR	http://creativecommons.org/licenses/by/4.0/	In this paper, we present a simulation and control framework for generating biomechanically plausible motion for muscle-actuated characters. We incorporate a fatigue dynamics model, the 3CC-r model, into the widely-adopted Hill-type muscle model to simulate the development and recovery of fatigue in muscles, which creates a natural evolution of motion style caused by the accumulation of fatigue from prolonged activities. To address the challenging problem of controlling a musculoskeletal system with high degrees of freedom, we propose a novel muscle-space control strategy based on PD control. Our simulation and control framework facilitates the training of a generative model for muscle-based motion control, which we refer to as MuscleVAE. By leveraging the variational autoencoders (VAEs), MuscleVAE is capable of learning a rich and flexible latent representation of skills from a large unstructured motion dataset, encoding not only motion features but also muscle control and fatigue properties. We demonstrate that the MuscleVAE model can be efficiently trained using a model-based approach, resulting in the production of high-fidelity motions and enabling a variety of downstream tasks.	[ { "created": "Tue, 12 Dec 2023 15:01:17 GMT", "version": "v1" } ]	2023-12-13	[ [ "Feng", "Yusen", "" ], [ "Xu", "Xiyan", "" ], [ "Liu", "Libin", "" ] ]	In this paper, we present a simulation and control framework for generating biomechanically plausible motion for muscle-actuated characters. We incorporate a fatigue dynamics model, the 3CC-r model, into the widely-adopted Hill-type muscle model to simulate the development and recovery of fatigue in muscles, which creates a natural evolution of motion style caused by the accumulation of fatigue from prolonged activities. To address the challenging problem of controlling a musculoskeletal system with high degrees of freedom, we propose a novel muscle-space control strategy based on PD control. Our simulation and control framework facilitates the training of a generative model for muscle-based motion control, which we refer to as MuscleVAE. By leveraging the variational autoencoders (VAEs), MuscleVAE is capable of learning a rich and flexible latent representation of skills from a large unstructured motion dataset, encoding not only motion features but also muscle control and fatigue properties. We demonstrate that the MuscleVAE model can be efficiently trained using a model-based approach, resulting in the production of high-fidelity motions and enabling a variety of downstream tasks.
2211.08292	Jennifer Andreoli-Fang	Jennifer Andreoli-Fang, John T Chapman	Mobile-Aware Scheduling for Low Latency Backhaul over DOCSIS	IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 2017	null	10.1109/PIMRC.2017.8292173	null	cs.NI	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	In this paper, we discuss latency reduction techniques for mobile backhaul over Data Over Cable Service Interface Specifications (DOCSIS) networks. When the latencies from both the wireless and the DOCSIS networks are added together, it can result in noticeable end-to-end system latency, particularly under network congestion. Previously, we proposed a method to improve upstream user-to-mobile core latency by coordinating the LTE and DOCSIS scheduling. The method reduces the impact on system latency from the DOCSIS network's request-grant-data loop, which is the main contributor of backhaul upstream latency. Since the method reduces latency on the DOCSIS data path, it will therefore improve performance of latency sensitive applications, particularly if TCP is used as the transport protocol, especially when the link is congested. In this paper, we investigate the effect of HARQ failure on system performance. Through simulation, we show that despite the uncertainty introduced by the LTE protocol, coordinated scheduling improves overall system latency.	[ { "created": "Tue, 15 Nov 2022 16:45:26 GMT", "version": "v1" }, { "created": "Wed, 16 Nov 2022 16:38:44 GMT", "version": "v2" } ]	2022-11-17	[ [ "Andreoli-Fang", "Jennifer", "" ], [ "Chapman", "John T", "" ] ]	In this paper, we discuss latency reduction techniques for mobile backhaul over Data Over Cable Service Interface Specifications (DOCSIS) networks. When the latencies from both the wireless and the DOCSIS networks are added together, it can result in noticeable end-to-end system latency, particularly under network congestion. Previously, we proposed a method to improve upstream user-to-mobile core latency by coordinating the LTE and DOCSIS scheduling. The method reduces the impact on system latency from the DOCSIS network's request-grant-data loop, which is the main contributor of backhaul upstream latency. Since the method reduces latency on the DOCSIS data path, it will therefore improve performance of latency sensitive applications, particularly if TCP is used as the transport protocol, especially when the link is congested. In this paper, we investigate the effect of HARQ failure on system performance. Through simulation, we show that despite the uncertainty introduced by the LTE protocol, coordinated scheduling improves overall system latency.
2103.14856	Ciriaco Andrea D'Angelo	Giovanni Abramo, Ciriaco Andrea D'Angelo, Lin Zhang	A comparison of two approaches for measuring interdisciplinary research output: the disciplinary diversity of authors vs the disciplinary diversity of the reference list	null	Journal of Informetrics, 12(4), 2018, 1182-1193	10.1016/j.joi.2018.09.001	null	cs.DL	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	This study investigates the convergence of two bibliometric approaches to the measurement of interdisciplinary research: one based on analyzing disciplinary diversity in the reference list of publications, the other based on the disciplinary diversity of authors of publications. In particular we measure the variety, balance, disparity and integrated diversity index of, respectively, single-author, multi-author single-field, and multi-author multi-field publications. We find that, in general, the diversity of the reference list grows with the number of fields reflected in a paper's authors' list and, to a lesser extent, with the number of authors being equal the number of fields. Further, we find that when fields belonging to different disciplines are reflected in the authors' list, the disparity in the reference list is higher than in the case of fields belonging to the same discipline. However, this general tendency varies across disciplines, and noticeable exceptions are found at individual paper level.	[ { "created": "Sat, 27 Mar 2021 09:34:53 GMT", "version": "v1" } ]	2021-03-30	[ [ "Abramo", "Giovanni", "" ], [ "D'Angelo", "Ciriaco Andrea", "" ], [ "Zhang", "Lin", "" ] ]	This study investigates the convergence of two bibliometric approaches to the measurement of interdisciplinary research: one based on analyzing disciplinary diversity in the reference list of publications, the other based on the disciplinary diversity of authors of publications. In particular we measure the variety, balance, disparity and integrated diversity index of, respectively, single-author, multi-author single-field, and multi-author multi-field publications. We find that, in general, the diversity of the reference list grows with the number of fields reflected in a paper's authors' list and, to a lesser extent, with the number of authors being equal the number of fields. Further, we find that when fields belonging to different disciplines are reflected in the authors' list, the disparity in the reference list is higher than in the case of fields belonging to the same discipline. However, this general tendency varies across disciplines, and noticeable exceptions are found at individual paper level.
1910.12435	Marcel Keller	Anders Dalskov and Daniel Escudero and Marcel Keller	Secure Evaluation of Quantized Neural Networks	22 pages	Proceedings on Privacy Enhancing Technologies 4 (2020): 355-375	10.2478/popets-2020-0077	null	cs.CR cs.LG	http://creativecommons.org/licenses/by-nc-nd/4.0/	We investigate two questions in this paper: First, we ask to what extent "MPC friendly" models are already supported by major Machine Learning frameworks such as TensorFlow or PyTorch. Prior works provide protocols that only work on fixed-point integers and specialized activation functions, two aspects that are not supported by popular Machine Learning frameworks, and the need for these specialized model representations means that it is hard, and often impossible, to use e.g., TensorFlow to design, train and test models that later have to be evaluated securely. Second, we ask to what extent the functionality for evaluating Neural Networks already exists in general-purpose MPC frameworks. These frameworks have received more scrutiny, are better documented and supported on more platforms. Furthermore, they are typically flexible in terms of the threat model they support. In contrast, most secure evaluation protocols in the literature are targeted to a specific threat model and their implementations are only a "proof-of-concept", making it very hard for their adoption in practice. We answer both of the above questions in a positive way: We observe that the quantization techniques supported by both TensorFlow, PyTorch and MXNet can provide models in a representation that can be evaluated securely; and moreover, that this evaluation can be performed by a general purpose MPC framework. We perform extensive benchmarks to understand the exact trade-offs between different corruption models, network sizes and efficiency. These experiments provide an interesting insight into cost between active and passive security, as well as honest and dishonest majority. Our work shows then that the separating line between existing ML frameworks and existing MPC protocols may be narrower than implicitly suggested by previous works.	[ { "created": "Mon, 28 Oct 2019 04:17:33 GMT", "version": "v1" }, { "created": "Mon, 1 Mar 2021 04:22:07 GMT", "version": "v2" } ]	2021-03-02	[ [ "Dalskov", "Anders", "" ], [ "Escudero", "Daniel", "" ], [ "Keller", "Marcel", "" ] ]	We investigate two questions in this paper: First, we ask to what extent "MPC friendly" models are already supported by major Machine Learning frameworks such as TensorFlow or PyTorch. Prior works provide protocols that only work on fixed-point integers and specialized activation functions, two aspects that are not supported by popular Machine Learning frameworks, and the need for these specialized model representations means that it is hard, and often impossible, to use e.g., TensorFlow to design, train and test models that later have to be evaluated securely. Second, we ask to what extent the functionality for evaluating Neural Networks already exists in general-purpose MPC frameworks. These frameworks have received more scrutiny, are better documented and supported on more platforms. Furthermore, they are typically flexible in terms of the threat model they support. In contrast, most secure evaluation protocols in the literature are targeted to a specific threat model and their implementations are only a "proof-of-concept", making it very hard for their adoption in practice. We answer both of the above questions in a positive way: We observe that the quantization techniques supported by both TensorFlow, PyTorch and MXNet can provide models in a representation that can be evaluated securely; and moreover, that this evaluation can be performed by a general purpose MPC framework. We perform extensive benchmarks to understand the exact trade-offs between different corruption models, network sizes and efficiency. These experiments provide an interesting insight into cost between active and passive security, as well as honest and dishonest majority. Our work shows then that the separating line between existing ML frameworks and existing MPC protocols may be narrower than implicitly suggested by previous works.
2407.06855	Arnab Sharma	Sourabh Kapoor, Arnab Sharma, Michael R\"oder, Caglar Demir, Axel-Cyrille Ngonga Ngomo	Performance Evaluation of Knowledge Graph Embedding Approaches under Non-adversarial Attacks	null	null	null	null	cs.LG cs.CR	http://creativecommons.org/licenses/by/4.0/	Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.	[ { "created": "Tue, 9 Jul 2024 13:42:14 GMT", "version": "v1" } ]	2024-07-10	[ [ "Kapoor", "Sourabh", "" ], [ "Sharma", "Arnab", "" ], [ "Röder", "Michael", "" ], [ "Demir", "Caglar", "" ], [ "Ngomo", "Axel-Cyrille Ngonga", "" ] ]	Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.
1211.1861	Mohamed Firdhous	Mohamed Firdhous	Automating Legal Research through Data Mining	8 pages, 11 figures, published in (IJACSA) International Journal of Advanced Computer Science and Applications. arXiv admin note: text overlap with wikipedia entry on text mining	International Journal of Advanced Computer Science and Applications (IJACSA) Vol. 1, No 6, December 2010, pp. 9-16	null	null	cs.IR	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	The term legal research generally refers to the process of identifying and retrieving appropriate information necessary to support legal decision making from past case records. At present, the process is mostly manual, but some traditional technologies such as keyword searching are commonly used to speed the process up. But a keyword search is not a comprehensive search to cater to the requirements of legal research as the search result includes too many false hits in terms of irrelevant case records. Hence the present generic tools cannot be used to automate legal research. This paper presents a framework which was developed by combining several Text Mining techniques to automate the process overcoming the difficulties in the existing methods. Further, the research also identifies the possible enhancements that could be done to enhance the effectiveness of the framework.	[ { "created": "Thu, 8 Nov 2012 14:19:49 GMT", "version": "v1" } ]	2012-11-09	[ [ "Firdhous", "Mohamed", "" ] ]	The term legal research generally refers to the process of identifying and retrieving appropriate information necessary to support legal decision making from past case records. At present, the process is mostly manual, but some traditional technologies such as keyword searching are commonly used to speed the process up. But a keyword search is not a comprehensive search to cater to the requirements of legal research as the search result includes too many false hits in terms of irrelevant case records. Hence the present generic tools cannot be used to automate legal research. This paper presents a framework which was developed by combining several Text Mining techniques to automate the process overcoming the difficulties in the existing methods. Further, the research also identifies the possible enhancements that could be done to enhance the effectiveness of the framework.
1602.00810	Jean-Guillaume Dumas	Jean-Guillaume Dumas (LJK), Erich Kaltofen (NCSU), Emmanuel Thom\'e (CARAMBA), Gilles Villard (ARIC, LIP)	Linear Time Interactive Certificates for the Minimal Polynomial and the Determinant of a Sparse Matrix	null	International Symposium on Symbolic and Algebraic Computation, Jul 2016, Waterloo, Canada. pp.199-206, \&\#x27E8;10.1145/2930889.2930908\&\#x27E9	null	null	cs.SC	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Computational problem certificates are additional data structures for each output, which can be used by a-possibly randomized-verification algorithm that proves the correctness of each output. In this paper, we give an algorithm that computes a certificate for the minimal polynomial of sparse or structured nxn matrices over an abstract field, of sufficiently large cardinality, whose Monte Carlo verification complexity requires a single matrix-vector multiplication and a linear number of extra field operations. We also propose a novel preconditioner that ensures irreducibility of the characteristic polynomial of the generically preconditioned matrix. This preconditioner takes linear time to be applied and uses only two random entries. We then combine these two techniques to give algorithms that compute certificates for the determinant, and thus for the characteristic polynomial, whose Monte Carlo verification complexity is therefore also linear.	[ { "created": "Tue, 2 Feb 2016 07:29:28 GMT", "version": "v1" }, { "created": "Mon, 2 Dec 2019 13:02:25 GMT", "version": "v2" } ]	2019-12-03	[ [ "Dumas", "Jean-Guillaume", "", "LJK" ], [ "Kaltofen", "Erich", "", "NCSU" ], [ "Thomé", "Emmanuel", "", "CARAMBA" ], [ "Villard", "Gilles", "", "ARIC, LIP" ] ]	Computational problem certificates are additional data structures for each output, which can be used by a-possibly randomized-verification algorithm that proves the correctness of each output. In this paper, we give an algorithm that computes a certificate for the minimal polynomial of sparse or structured nxn matrices over an abstract field, of sufficiently large cardinality, whose Monte Carlo verification complexity requires a single matrix-vector multiplication and a linear number of extra field operations. We also propose a novel preconditioner that ensures irreducibility of the characteristic polynomial of the generically preconditioned matrix. This preconditioner takes linear time to be applied and uses only two random entries. We then combine these two techniques to give algorithms that compute certificates for the determinant, and thus for the characteristic polynomial, whose Monte Carlo verification complexity is therefore also linear.
1912.07976	Heng Yang	Heng Yang, Biqing Zeng, JianHao Yang, Youwei Song and Ruyang Xu	A Multi-task Learning Model for Chinese-oriented Aspect Polarity Classification and Aspect Term Extraction	Submitted to Elsevier	null	null	null	cs.CL cs.LG	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Aspect-based sentiment analysis (ABSA) task is a multi-grained task of natural language processing and consists of two subtasks: aspect term extraction (ATE) and aspect polarity classification (APC). Most of the existing work focuses on the subtask of aspect term polarity inferring and ignores the significance of aspect term extraction. Besides, the existing researches do not pay attention to the research of the Chinese-oriented ABSA task. Based on the local context focus (LCF) mechanism, this paper firstly proposes a multi-task learning model for Chinese-oriented aspect-based sentiment analysis, namely LCF-ATEPC. Compared with existing models, this model equips the capability of extracting aspect term and inferring aspect term polarity synchronously, moreover, this model is effective to analyze both Chinese and English comments simultaneously and the experiment on a multilingual mixed dataset proved its availability. By integrating the domain-adapted BERT model, the LCF-ATEPC model achieved the state-of-the-art performance of aspect term extraction and aspect polarity classification in four Chinese review datasets. Besides, the experimental results on the most commonly used SemEval-2014 task4 Restaurant and Laptop datasets outperform the state-of-the-art performance on the ATE and APC subtask.	[ { "created": "Tue, 17 Dec 2019 12:47:33 GMT", "version": "v1" }, { "created": "Thu, 19 Dec 2019 01:38:38 GMT", "version": "v2" }, { "created": "Wed, 12 Feb 2020 09:20:28 GMT", "version": "v3" } ]	2020-02-13	[ [ "Yang", "Heng", "" ], [ "Zeng", "Biqing", "" ], [ "Yang", "JianHao", "" ], [ "Song", "Youwei", "" ], [ "Xu", "Ruyang", "" ] ]	Aspect-based sentiment analysis (ABSA) task is a multi-grained task of natural language processing and consists of two subtasks: aspect term extraction (ATE) and aspect polarity classification (APC). Most of the existing work focuses on the subtask of aspect term polarity inferring and ignores the significance of aspect term extraction. Besides, the existing researches do not pay attention to the research of the Chinese-oriented ABSA task. Based on the local context focus (LCF) mechanism, this paper firstly proposes a multi-task learning model for Chinese-oriented aspect-based sentiment analysis, namely LCF-ATEPC. Compared with existing models, this model equips the capability of extracting aspect term and inferring aspect term polarity synchronously, moreover, this model is effective to analyze both Chinese and English comments simultaneously and the experiment on a multilingual mixed dataset proved its availability. By integrating the domain-adapted BERT model, the LCF-ATEPC model achieved the state-of-the-art performance of aspect term extraction and aspect polarity classification in four Chinese review datasets. Besides, the experimental results on the most commonly used SemEval-2014 task4 Restaurant and Laptop datasets outperform the state-of-the-art performance on the ATE and APC subtask.
2007.00584	Guillaume Jaume	Pushpak Pati, Guillaume Jaume, Lauren Alisha Fernandes, Antonio Foncubierta, Florinda Feroce, Anna Maria Anniciello, Giosue Scognamiglio, Nadia Brancati, Daniel Riccio, Maurizio Do Bonito, Giuseppe De Pietro, Gerardo Botti, Orcun Goksel, Jean-Philippe Thiran, Maria Frucci, Maria Gabrani	HACT-Net: A Hierarchical Cell-to-Tissue Graph Neural Network for Histopathological Image Classification	null	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Cancer diagnosis, prognosis, and therapeutic response prediction are heavily influenced by the relationship between the histopathological structures and the function of the tissue. Recent approaches acknowledging the structure-function relationship, have linked the structural and spatial patterns of cell organization in tissue via cell-graphs to tumor grades. Though cell organization is imperative, it is insufficient to entirely represent the histopathological structure. We propose a novel hierarchical cell-to-tissue-graph (HACT) representation to improve the structural depiction of the tissue. It consists of a low-level cell-graph, capturing cell morphology and interactions, a high-level tissue-graph, capturing morphology and spatial distribution of tissue parts, and cells-to-tissue hierarchies, encoding the relative spatial distribution of the cells with respect to the tissue distribution. Further, a hierarchical graph neural network (HACT-Net) is proposed to efficiently map the HACT representations to histopathological breast cancer subtypes. We assess the methodology on a large set of annotated tissue regions of interest from H\&E stained breast carcinoma whole-slides. Upon evaluation, the proposed method outperformed recent convolutional neural network and graph neural network approaches for breast cancer multi-class subtyping. The proposed entity-based topological analysis is more inline with the pathological diagnostic procedure of the tissue. It provides more command over the tissue modelling, therefore encourages the further inclusion of pathological priors into task-specific tissue representation.	[ { "created": "Wed, 1 Jul 2020 16:22:48 GMT", "version": "v1" } ]	2020-07-02	[ [ "Pati", "Pushpak", "" ], [ "Jaume", "Guillaume", "" ], [ "Fernandes", "Lauren Alisha", "" ], [ "Foncubierta", "Antonio", "" ], [ "Feroce", "Florinda", "" ], [ "Anniciello", "Anna Maria", "" ], [ "Scognamiglio", "Giosue", "" ], [ "Brancati", "Nadia", "" ], [ "Riccio", "Daniel", "" ], [ "Bonito", "Maurizio Do", "" ], [ "De Pietro", "Giuseppe", "" ], [ "Botti", "Gerardo", "" ], [ "Goksel", "Orcun", "" ], [ "Thiran", "Jean-Philippe", "" ], [ "Frucci", "Maria", "" ], [ "Gabrani", "Maria", "" ] ]	Cancer diagnosis, prognosis, and therapeutic response prediction are heavily influenced by the relationship between the histopathological structures and the function of the tissue. Recent approaches acknowledging the structure-function relationship, have linked the structural and spatial patterns of cell organization in tissue via cell-graphs to tumor grades. Though cell organization is imperative, it is insufficient to entirely represent the histopathological structure. We propose a novel hierarchical cell-to-tissue-graph (HACT) representation to improve the structural depiction of the tissue. It consists of a low-level cell-graph, capturing cell morphology and interactions, a high-level tissue-graph, capturing morphology and spatial distribution of tissue parts, and cells-to-tissue hierarchies, encoding the relative spatial distribution of the cells with respect to the tissue distribution. Further, a hierarchical graph neural network (HACT-Net) is proposed to efficiently map the HACT representations to histopathological breast cancer subtypes. We assess the methodology on a large set of annotated tissue regions of interest from H\&E stained breast carcinoma whole-slides. Upon evaluation, the proposed method outperformed recent convolutional neural network and graph neural network approaches for breast cancer multi-class subtyping. The proposed entity-based topological analysis is more inline with the pathological diagnostic procedure of the tissue. It provides more command over the tissue modelling, therefore encourages the further inclusion of pathological priors into task-specific tissue representation.
0810.5057	Patricia Gautier	Claire Fran\c{c}ois (INIST), Jean-Charles Lamirel (INRIA Lorraine - LORIA), Shadi Al Shehabi (INRIA Lorraine - LORIA)	Combining Advanced Visualization and Automatized Reasoning for Webometrics: A Test Study	null	COLLNET 2006, France (2006)	null	null	cs.IR cs.DL	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	This paper presents a first attempt at performing a precise and automatic identification of the linking behaviour in a scientific domain through the analysis of the communication of the related academic institutions on the web. The proposed approach is based on the paradigm of multiple viewpoint data analysis (MVDA) than can be fruitfully exploited to highlight relationships between data, like websites, carrying several kinds of description. It uses the MultiSOM clustering and mapping method. The domain that has been chosen for this study is the domain of Computer Science in Germany. The analysis is conduced on a set of 438 websites of this domain using all together, thematic, geographic and linking information. It highlights interesting results concerning both global and local linking behaviour.	[ { "created": "Tue, 28 Oct 2008 15:43:45 GMT", "version": "v1" }, { "created": "Tue, 20 Oct 2009 15:27:41 GMT", "version": "v2" } ]	2009-10-20	[ [ "François", "Claire", "", "INIST" ], [ "Lamirel", "Jean-Charles", "", "INRIA Lorraine -\n LORIA" ], [ "Shehabi", "Shadi Al", "", "INRIA Lorraine - LORIA" ] ]	This paper presents a first attempt at performing a precise and automatic identification of the linking behaviour in a scientific domain through the analysis of the communication of the related academic institutions on the web. The proposed approach is based on the paradigm of multiple viewpoint data analysis (MVDA) than can be fruitfully exploited to highlight relationships between data, like websites, carrying several kinds of description. It uses the MultiSOM clustering and mapping method. The domain that has been chosen for this study is the domain of Computer Science in Germany. The analysis is conduced on a set of 438 websites of this domain using all together, thematic, geographic and linking information. It highlights interesting results concerning both global and local linking behaviour.
0808.3651	Lijun Zhang	Lijun Zhang, Holger Hermanns, Friedrich Eisenbrand and David N. Jansen	Flow Faster: Efficient Decision Algorithms for Probabilistic Simulations	LMCS	Logical Methods in Computer Science, Volume 4, Issue 4 (November 11, 2008) lmcs:989	10.2168/LMCS-4(4:6)2008	null	cs.LO	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Strong and weak simulation relations have been proposed for Markov chains, while strong simulation and strong probabilistic simulation relations have been proposed for probabilistic automata. However, decision algorithms for strong and weak simulation over Markov chains, and for strong simulation over probabilistic automata are not efficient, which makes it as yet unclear whether they can be used as effectively as their non-probabilistic counterparts. This paper presents drastically improved algorithms to decide whether some (discrete- or continuous-time) Markov chain strongly or weakly simulates another, or whether a probabilistic automaton strongly simulates another. The key innovation is the use of parametric maximum flow techniques to amortize computations. We also present a novel algorithm for deciding strong probabilistic simulation preorders on probabilistic automata, which has polynomial complexity via a reduction to an LP problem. When extending the algorithms for probabilistic automata to their continuous-time counterpart, we retain the same complexity for both strong and strong probabilistic simulations.	[ { "created": "Wed, 27 Aug 2008 08:35:44 GMT", "version": "v1" }, { "created": "Mon, 10 Nov 2008 23:56:01 GMT", "version": "v2" }, { "created": "Tue, 18 Nov 2008 17:00:30 GMT", "version": "v3" } ]	2015-07-01	[ [ "Zhang", "Lijun", "" ], [ "Hermanns", "Holger", "" ], [ "Eisenbrand", "Friedrich", "" ], [ "Jansen", "David N.", "" ] ]	Strong and weak simulation relations have been proposed for Markov chains, while strong simulation and strong probabilistic simulation relations have been proposed for probabilistic automata. However, decision algorithms for strong and weak simulation over Markov chains, and for strong simulation over probabilistic automata are not efficient, which makes it as yet unclear whether they can be used as effectively as their non-probabilistic counterparts. This paper presents drastically improved algorithms to decide whether some (discrete- or continuous-time) Markov chain strongly or weakly simulates another, or whether a probabilistic automaton strongly simulates another. The key innovation is the use of parametric maximum flow techniques to amortize computations. We also present a novel algorithm for deciding strong probabilistic simulation preorders on probabilistic automata, which has polynomial complexity via a reduction to an LP problem. When extending the algorithms for probabilistic automata to their continuous-time counterpart, we retain the same complexity for both strong and strong probabilistic simulations.
2403.06757	Anthony Frion	Anthony Frion, Lucas Drumetz, Guillaume Tochon, Mauro Dalla Mura, Albdeldjalil A\"issa El Bey	Koopman Ensembles for Probabilistic Time Series Forecasting	null	null	null	null	cs.LG	http://creativecommons.org/licenses/by/4.0/	In the context of an increasing popularity of data-driven models to represent dynamical systems, many machine learning-based implementations of the Koopman operator have recently been proposed. However, the vast majority of those works are limited to deterministic predictions, while the knowledge of uncertainty is critical in fields like meteorology and climatology. In this work, we investigate the training of ensembles of models to produce stochastic outputs. We show through experiments on real remote sensing image time series that ensembles of independently trained models are highly overconfident and that using a training criterion that explicitly encourages the members to produce predictions with high inter-model variances greatly improves the uncertainty quantification of the ensembles.	[ { "created": "Mon, 11 Mar 2024 14:29:56 GMT", "version": "v1" }, { "created": "Wed, 13 Mar 2024 13:57:42 GMT", "version": "v2" } ]	2024-03-14	[ [ "Frion", "Anthony", "" ], [ "Drumetz", "Lucas", "" ], [ "Tochon", "Guillaume", "" ], [ "Mura", "Mauro Dalla", "" ], [ "Bey", "Albdeldjalil Aïssa El", "" ] ]	In the context of an increasing popularity of data-driven models to represent dynamical systems, many machine learning-based implementations of the Koopman operator have recently been proposed. However, the vast majority of those works are limited to deterministic predictions, while the knowledge of uncertainty is critical in fields like meteorology and climatology. In this work, we investigate the training of ensembles of models to produce stochastic outputs. We show through experiments on real remote sensing image time series that ensembles of independently trained models are highly overconfident and that using a training criterion that explicitly encourages the members to produce predictions with high inter-model variances greatly improves the uncertainty quantification of the ensembles.
1702.06968	Arthur-Jozsef Molnar	Arthur-Jozsef Molnar	A Heuristic Process for GUI Widget Matching Across Application Versions	null	Annales Universitatis Scientiarum Budapest, Sectio Computatorica Vol 36, pages 255 - 275 (2012)	null	null	cs.SE	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	This paper introduces an automated heuristic process able to achieve high accuracy when matching graphical user interface widgets across multiple versions of a target application. The proposed implementation is flexible as it allows full customization of the process and easy integration with existing tools for long term graphical user interface test case maintenance, software visualization and analysis.	[ { "created": "Wed, 22 Feb 2017 19:08:11 GMT", "version": "v1" } ]	2017-02-24	[ [ "Molnar", "Arthur-Jozsef", "" ] ]	This paper introduces an automated heuristic process able to achieve high accuracy when matching graphical user interface widgets across multiple versions of a target application. The proposed implementation is flexible as it allows full customization of the process and easy integration with existing tools for long term graphical user interface test case maintenance, software visualization and analysis.
2307.10936	Raphael Boige	Raphael Boige and Yannis Flet-Berliac and Arthur Flajolet and Guillaume Richard and Thomas Pierrot	PASTA: Pretrained Action-State Transformer Agents	null	null	null	null	cs.AI cs.LG	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Self-supervised learning has brought about a revolutionary paradigm shift in various computing domains, including NLP, vision, and biology. Recent approaches involve pre-training transformer models on vast amounts of unlabeled data, serving as a starting point for efficiently solving downstream tasks. In reinforcement learning, researchers have recently adapted these approaches, developing models pre-trained on expert trajectories. This advancement enables the models to tackle a broad spectrum of tasks, ranging from robotics to recommendation systems. However, existing methods mostly rely on intricate pre-training objectives tailored to specific downstream applications. This paper conducts a comprehensive investigation of models, referred to as pre-trained action-state transformer agents (PASTA). Our study covers a unified methodology and covers an extensive set of general downstream tasks including behavioral cloning, offline RL, sensor failure robustness, and dynamics change adaptation. Our objective is to systematically compare various design choices and offer valuable insights that will aid practitioners in developing robust models. Key highlights of our study include tokenization at the component level for actions and states, the use of fundamental pre-training objectives such as next token prediction or masked language modeling, simultaneous training of models across multiple domains, and the application of various fine-tuning strategies. In this study, the developed models contain fewer than 7 million parameters allowing a broad community to use these models and reproduce our experiments. We hope that this study will encourage further research into the use of transformers with first principle design choices to represent RL trajectories and contribute to robust policy learning.	[ { "created": "Thu, 20 Jul 2023 15:09:06 GMT", "version": "v1" }, { "created": "Mon, 4 Dec 2023 10:15:26 GMT", "version": "v2" } ]	2023-12-05	[ [ "Boige", "Raphael", "" ], [ "Flet-Berliac", "Yannis", "" ], [ "Flajolet", "Arthur", "" ], [ "Richard", "Guillaume", "" ], [ "Pierrot", "Thomas", "" ] ]	Self-supervised learning has brought about a revolutionary paradigm shift in various computing domains, including NLP, vision, and biology. Recent approaches involve pre-training transformer models on vast amounts of unlabeled data, serving as a starting point for efficiently solving downstream tasks. In reinforcement learning, researchers have recently adapted these approaches, developing models pre-trained on expert trajectories. This advancement enables the models to tackle a broad spectrum of tasks, ranging from robotics to recommendation systems. However, existing methods mostly rely on intricate pre-training objectives tailored to specific downstream applications. This paper conducts a comprehensive investigation of models, referred to as pre-trained action-state transformer agents (PASTA). Our study covers a unified methodology and covers an extensive set of general downstream tasks including behavioral cloning, offline RL, sensor failure robustness, and dynamics change adaptation. Our objective is to systematically compare various design choices and offer valuable insights that will aid practitioners in developing robust models. Key highlights of our study include tokenization at the component level for actions and states, the use of fundamental pre-training objectives such as next token prediction or masked language modeling, simultaneous training of models across multiple domains, and the application of various fine-tuning strategies. In this study, the developed models contain fewer than 7 million parameters allowing a broad community to use these models and reproduce our experiments. We hope that this study will encourage further research into the use of transformers with first principle design choices to represent RL trajectories and contribute to robust policy learning.
1612.00604	Andrii Maksai	Andrii Maksai, Xinchao Wang, Francois Fleuret, and Pascal Fua	Globally Consistent Multi-People Tracking using Motion Patterns	8 pages, 7 figures. 11 pages supplementary	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Many state-of-the-art approaches to people tracking rely on detecting them in each frame independently, grouping detections into short but reliable trajectory segments, and then further grouping them into full trajectories. This grouping typically relies on imposing local smoothness constraints but almost never on enforcing more global constraints on the trajectories. In this paper, we propose an approach to imposing global consistency by first inferring behavioral patterns from the ground truth and then using them to guide the tracking algorithm. When used in conjunction with several state-of-the-art algorithms, this further increases their already good performance. Furthermore, we propose an unsupervised scheme that yields almost similar improvements without the need for ground truth.	[ { "created": "Fri, 2 Dec 2016 09:24:30 GMT", "version": "v1" } ]	2016-12-05	[ [ "Maksai", "Andrii", "" ], [ "Wang", "Xinchao", "" ], [ "Fleuret", "Francois", "" ], [ "Fua", "Pascal", "" ] ]	Many state-of-the-art approaches to people tracking rely on detecting them in each frame independently, grouping detections into short but reliable trajectory segments, and then further grouping them into full trajectories. This grouping typically relies on imposing local smoothness constraints but almost never on enforcing more global constraints on the trajectories. In this paper, we propose an approach to imposing global consistency by first inferring behavioral patterns from the ground truth and then using them to guide the tracking algorithm. When used in conjunction with several state-of-the-art algorithms, this further increases their already good performance. Furthermore, we propose an unsupervised scheme that yields almost similar improvements without the need for ground truth.
2311.11532	Gustavo Silva	Gustavo Silva, Paul Rodriguez	Optimal Hyperparameter $\epsilon$ for Adaptive Stochastic Optimizers through Gradient Histograms	null	null	null	null	cs.LG cs.AI	http://creativecommons.org/licenses/by-nc-nd/4.0/	Optimizers are essential components for successfully training deep neural network models. In order to achieve the best performance from such models, designers need to carefully choose the optimizer hyperparameters. However, this can be a computationally expensive and time-consuming process. Although it is known that all optimizer hyperparameters must be tuned for maximum performance, there is still a lack of clarity regarding the individual influence of minor priority hyperparameters, including the safeguard factor $\epsilon$ and momentum factor $\beta$, in leading adaptive optimizers (specifically, those based on the Adam optimizers). In this manuscript, we introduce a new framework based on gradient histograms to analyze and justify important attributes of adaptive optimizers, such as their optimal performance and the relationships and dependencies among hyperparameters. Furthermore, we propose a novel gradient histogram-based algorithm that automatically estimates a reduced and accurate search space for the safeguard hyperparameter $\epsilon$, where the optimal value can be easily found.	[ { "created": "Mon, 20 Nov 2023 04:34:19 GMT", "version": "v1" } ]	2023-11-21	[ [ "Silva", "Gustavo", "" ], [ "Rodriguez", "Paul", "" ] ]	Optimizers are essential components for successfully training deep neural network models. In order to achieve the best performance from such models, designers need to carefully choose the optimizer hyperparameters. However, this can be a computationally expensive and time-consuming process. Although it is known that all optimizer hyperparameters must be tuned for maximum performance, there is still a lack of clarity regarding the individual influence of minor priority hyperparameters, including the safeguard factor $\epsilon$ and momentum factor $\beta$, in leading adaptive optimizers (specifically, those based on the Adam optimizers). In this manuscript, we introduce a new framework based on gradient histograms to analyze and justify important attributes of adaptive optimizers, such as their optimal performance and the relationships and dependencies among hyperparameters. Furthermore, we propose a novel gradient histogram-based algorithm that automatically estimates a reduced and accurate search space for the safeguard hyperparameter $\epsilon$, where the optimal value can be easily found.
2302.02187	Niklas K\"uhl Prof Dr	Max Schemmer, Niklas K\"uhl, Carina Benz, Andrea Bartos, Gerhard Satzger	Appropriate Reliance on AI Advice: Conceptualization and the Effect of Explanations	arXiv admin note: text overlap with arXiv:2204.06916	ACM 28th International Conference on Intelligent User Interfaces (IUI), 2023	10.1145/3581641.3584066	null	cs.AI cs.HC	http://creativecommons.org/licenses/by-nc-nd/4.0/	AI advice is becoming increasingly popular, e.g., in investment and medical treatment decisions. As this advice is typically imperfect, decision-makers have to exert discretion as to whether actually follow that advice: they have to "appropriately" rely on correct and turn down incorrect advice. However, current research on appropriate reliance still lacks a common definition as well as an operational measurement concept. Additionally, no in-depth behavioral experiments have been conducted that help understand the factors influencing this behavior. In this paper, we propose Appropriateness of Reliance (AoR) as an underlying, quantifiable two-dimensional measurement concept. We develop a research model that analyzes the effect of providing explanations for AI advice. In an experiment with 200 participants, we demonstrate how these explanations influence the AoR, and, thus, the effectiveness of AI advice. Our work contributes fundamental concepts for the analysis of reliance behavior and the purposeful design of AI advisors.	[ { "created": "Sat, 4 Feb 2023 15:48:24 GMT", "version": "v1" }, { "created": "Tue, 7 Feb 2023 07:47:43 GMT", "version": "v2" }, { "created": "Thu, 13 Apr 2023 08:50:16 GMT", "version": "v3" } ]	2023-04-14	[ [ "Schemmer", "Max", "" ], [ "Kühl", "Niklas", "" ], [ "Benz", "Carina", "" ], [ "Bartos", "Andrea", "" ], [ "Satzger", "Gerhard", "" ] ]	AI advice is becoming increasingly popular, e.g., in investment and medical treatment decisions. As this advice is typically imperfect, decision-makers have to exert discretion as to whether actually follow that advice: they have to "appropriately" rely on correct and turn down incorrect advice. However, current research on appropriate reliance still lacks a common definition as well as an operational measurement concept. Additionally, no in-depth behavioral experiments have been conducted that help understand the factors influencing this behavior. In this paper, we propose Appropriateness of Reliance (AoR) as an underlying, quantifiable two-dimensional measurement concept. We develop a research model that analyzes the effect of providing explanations for AI advice. In an experiment with 200 participants, we demonstrate how these explanations influence the AoR, and, thus, the effectiveness of AI advice. Our work contributes fundamental concepts for the analysis of reliance behavior and the purposeful design of AI advisors.
1101.1815	Suvansh Lal Mr.	Suvansh Lal, Mohit Jain, Vikrant Chaplot	Approaches to Formal Verification of Security Protocols	null	null	null	null	cs.CR	http://creativecommons.org/licenses/by/3.0/	In recent times, many protocols have been proposed to provide security for various information and communication systems. Such protocols must be tested for their functional correctness before they are used in practice. Application of formal methods for verification of security protocols would enhance their reliability thereby, increasing the usability of systems that employ them. Thus, formal verification of security protocols has become a key issue in computer and communications security. In this paper we present, analyze and compare some prevalent approaches towards verification of secure systems. We follow the notion of - same goal through different approaches - as we formally analyze the Needham Schroeder Public Key protocol for Lowe's attack using each of our presented approaches.	[ { "created": "Mon, 10 Jan 2011 13:53:25 GMT", "version": "v1" } ]	2011-01-11	[ [ "Lal", "Suvansh", "" ], [ "Jain", "Mohit", "" ], [ "Chaplot", "Vikrant", "" ] ]	In recent times, many protocols have been proposed to provide security for various information and communication systems. Such protocols must be tested for their functional correctness before they are used in practice. Application of formal methods for verification of security protocols would enhance their reliability thereby, increasing the usability of systems that employ them. Thus, formal verification of security protocols has become a key issue in computer and communications security. In this paper we present, analyze and compare some prevalent approaches towards verification of secure systems. We follow the notion of - same goal through different approaches - as we formally analyze the Needham Schroeder Public Key protocol for Lowe's attack using each of our presented approaches.
2405.11530	Sejik Park	Sejik Park	Learning More Generalized Experts by Merging Experts in Mixture-of-Experts	12 pages, 3 figures	null	null	null	cs.LG	http://creativecommons.org/licenses/by/4.0/	We observe that incorporating a shared layer in a mixture-of-experts can lead to performance degradation. This leads us to hypothesize that learning shared features poses challenges in deep learning, potentially caused by the same feature being learned as various different features. To address this issue, we track each expert's usage frequency and merge the two most frequently selected experts. We then update the least frequently selected expert using the combination of experts. This approach, combined with the subsequent learning of the router's expert selection, allows the model to determine if the most frequently selected experts have learned the same feature differently. If they have, the combined expert can be further trained to learn a more general feature. Consequently, our algorithm enhances transfer learning and mitigates catastrophic forgetting when applied to multi-domain task incremental learning.	[ { "created": "Sun, 19 May 2024 11:55:48 GMT", "version": "v1" } ]	2024-05-21	[ [ "Park", "Sejik", "" ] ]	We observe that incorporating a shared layer in a mixture-of-experts can lead to performance degradation. This leads us to hypothesize that learning shared features poses challenges in deep learning, potentially caused by the same feature being learned as various different features. To address this issue, we track each expert's usage frequency and merge the two most frequently selected experts. We then update the least frequently selected expert using the combination of experts. This approach, combined with the subsequent learning of the router's expert selection, allows the model to determine if the most frequently selected experts have learned the same feature differently. If they have, the combined expert can be further trained to learn a more general feature. Consequently, our algorithm enhances transfer learning and mitigates catastrophic forgetting when applied to multi-domain task incremental learning.
1701.01216	Tony T. Luo	T. Luo, S. S. Kanhere, H-P. Tan, F. Wu, and H. Wu	Crowdsourcing with Tullock contests: A new perspective	9 pages, 4 figures, 3 tables	Proc. IEEE INFOCOM, 2015, pp. 2515-2523	10.1109/INFOCOM.2015.7218641	null	cs.GT cs.HC cs.MA cs.NI	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Incentive mechanisms for crowdsourcing have been extensively studied under the framework of all-pay auctions. Along a distinct line, this paper proposes to use Tullock contests as an alternative tool to design incentive mechanisms for crowdsourcing. We are inspired by the conduciveness of Tullock contests to attracting user entry (yet not necessarily a higher revenue) in other domains. In this paper, we explore a new dimension in optimal Tullock contest design, by superseding the contest prize---which is fixed in conventional Tullock contests---with a prize function that is dependent on the (unknown) winner's contribution, in order to maximize the crowdsourcer's utility. We show that this approach leads to attractive practical advantages: (a) it is well-suited for rapid prototyping in fully distributed web agents and smartphone apps; (b) it overcomes the disincentive to participate caused by players' antagonism to an increasing number of rivals. Furthermore, we optimize conventional, fixed-prize Tullock contests to construct the most superior benchmark to compare against our mechanism. Through extensive evaluations, we show that our mechanism significantly outperforms the optimal benchmark, by over three folds on the crowdsourcer's utility cum profit and up to nine folds on the players' social welfare.	[ { "created": "Thu, 5 Jan 2017 05:44:25 GMT", "version": "v1" } ]	2017-01-06	[ [ "Luo", "T.", "" ], [ "Kanhere", "S. S.", "" ], [ "Tan", "H-P.", "" ], [ "Wu", "F.", "" ], [ "Wu", "H.", "" ] ]	Incentive mechanisms for crowdsourcing have been extensively studied under the framework of all-pay auctions. Along a distinct line, this paper proposes to use Tullock contests as an alternative tool to design incentive mechanisms for crowdsourcing. We are inspired by the conduciveness of Tullock contests to attracting user entry (yet not necessarily a higher revenue) in other domains. In this paper, we explore a new dimension in optimal Tullock contest design, by superseding the contest prize---which is fixed in conventional Tullock contests---with a prize function that is dependent on the (unknown) winner's contribution, in order to maximize the crowdsourcer's utility. We show that this approach leads to attractive practical advantages: (a) it is well-suited for rapid prototyping in fully distributed web agents and smartphone apps; (b) it overcomes the disincentive to participate caused by players' antagonism to an increasing number of rivals. Furthermore, we optimize conventional, fixed-prize Tullock contests to construct the most superior benchmark to compare against our mechanism. Through extensive evaluations, we show that our mechanism significantly outperforms the optimal benchmark, by over three folds on the crowdsourcer's utility cum profit and up to nine folds on the players' social welfare.
1704.00355	Neha Gupta	Moses Charikar, Neha Gupta, Roy Schwartz	Local Guarantees in Graph Cuts and Clustering	null	null	null	null	cs.DS	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Correlation Clustering is an elegant model that captures fundamental graph cut problems such as Min $s-t$ Cut, Multiway Cut, and Multicut, extensively studied in combinatorial optimization. Here, we are given a graph with edges labeled $+$ or $-$ and the goal is to produce a clustering that agrees with the labels as much as possible: $+$ edges within clusters and $-$ edges across clusters. The classical approach towards Correlation Clustering (and other graph cut problems) is to optimize a global objective. We depart from this and study local objectives: minimizing the maximum number of disagreements for edges incident on a single node, and the analogous max min agreements objective. This naturally gives rise to a family of basic min-max graph cut problems. A prototypical representative is Min Max $s-t$ Cut: find an $s-t$ cut minimizing the largest number of cut edges incident on any node. We present the following results: $(1)$ an $O(\sqrt{n})$-approximation for the problem of minimizing the maximum total weight of disagreement edges incident on any node (thus providing the first known approximation for the above family of min-max graph cut problems), $(2)$ a remarkably simple $7$-approximation for minimizing local disagreements in complete graphs (improving upon the previous best known approximation of $48$), and $(3)$ a $1/(2+\varepsilon)$-approximation for maximizing the minimum total weight of agreement edges incident on any node, hence improving upon the $1/(4+\varepsilon)$-approximation that follows from the study of approximate pure Nash equilibria in cut and party affiliation games.	[ { "created": "Sun, 2 Apr 2017 19:34:22 GMT", "version": "v1" } ]	2017-04-04	[ [ "Charikar", "Moses", "" ], [ "Gupta", "Neha", "" ], [ "Schwartz", "Roy", "" ] ]	Correlation Clustering is an elegant model that captures fundamental graph cut problems such as Min $s-t$ Cut, Multiway Cut, and Multicut, extensively studied in combinatorial optimization. Here, we are given a graph with edges labeled $+$ or $-$ and the goal is to produce a clustering that agrees with the labels as much as possible: $+$ edges within clusters and $-$ edges across clusters. The classical approach towards Correlation Clustering (and other graph cut problems) is to optimize a global objective. We depart from this and study local objectives: minimizing the maximum number of disagreements for edges incident on a single node, and the analogous max min agreements objective. This naturally gives rise to a family of basic min-max graph cut problems. A prototypical representative is Min Max $s-t$ Cut: find an $s-t$ cut minimizing the largest number of cut edges incident on any node. We present the following results: $(1)$ an $O(\sqrt{n})$-approximation for the problem of minimizing the maximum total weight of disagreement edges incident on any node (thus providing the first known approximation for the above family of min-max graph cut problems), $(2)$ a remarkably simple $7$-approximation for minimizing local disagreements in complete graphs (improving upon the previous best known approximation of $48$), and $(3)$ a $1/(2+\varepsilon)$-approximation for maximizing the minimum total weight of agreement edges incident on any node, hence improving upon the $1/(4+\varepsilon)$-approximation that follows from the study of approximate pure Nash equilibria in cut and party affiliation games.
1502.03358	Farshad Naghibi	Farshad Naghibi, Somayeh Salimi, Mikael Skoglund	The CEO Problem with Secrecy Constraints	Accepted for publication in IEEE Transactions on Information Forensics and Security, 17 pages, 4 figures	null	10.1109/TIFS.2015.2404134	null	cs.IT math.IT	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	We study a lossy source coding problem with secrecy constraints in which a remote information source should be transmitted to a single destination via multiple agents in the presence of a passive eavesdropper. The agents observe noisy versions of the source and independently encode and transmit their observations to the destination via noiseless rate-limited links. The destination should estimate the remote source based on the information received from the agents within a certain mean distortion threshold. The eavesdropper, with access to side information correlated to the source, is able to listen in on one of the links from the agents to the destination in order to obtain as much information as possible about the source. This problem can be viewed as the so-called CEO problem with additional secrecy constraints. We establish inner and outer bounds on the rate-distortion-equivocation region of this problem. We also obtain the region in special cases where the bounds are tight. Furthermore, we study the quadratic Gaussian case and provide the optimal rate-distortion-equivocation region when the eavesdropper has no side information and an achievable region for a more general setup with side information at the eavesdropper.	[ { "created": "Wed, 11 Feb 2015 16:19:21 GMT", "version": "v1" } ]	2015-02-19	[ [ "Naghibi", "Farshad", "" ], [ "Salimi", "Somayeh", "" ], [ "Skoglund", "Mikael", "" ] ]	We study a lossy source coding problem with secrecy constraints in which a remote information source should be transmitted to a single destination via multiple agents in the presence of a passive eavesdropper. The agents observe noisy versions of the source and independently encode and transmit their observations to the destination via noiseless rate-limited links. The destination should estimate the remote source based on the information received from the agents within a certain mean distortion threshold. The eavesdropper, with access to side information correlated to the source, is able to listen in on one of the links from the agents to the destination in order to obtain as much information as possible about the source. This problem can be viewed as the so-called CEO problem with additional secrecy constraints. We establish inner and outer bounds on the rate-distortion-equivocation region of this problem. We also obtain the region in special cases where the bounds are tight. Furthermore, we study the quadratic Gaussian case and provide the optimal rate-distortion-equivocation region when the eavesdropper has no side information and an achievable region for a more general setup with side information at the eavesdropper.
2007.00236	Cunxiang Wang	Cunxiang Wang, Shuailong Liang, Yili Jin, Yilong Wang, Xiaodan Zhu and Yue Zhang	SemEval-2020 Task 4: Commonsense Validation and Explanation	Task description paper of SemEval-2020 Task 4: Commonsense Validation and Explanation	null	null	null	cs.CL cs.AI	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	In this paper, we present SemEval-2020 Task 4, Commonsense Validation and Explanation (ComVE), which includes three subtasks, aiming to evaluate whether a system can distinguish a natural language statement that makes sense to humans from one that does not, and provide the reasons. Specifically, in our first subtask, the participating systems are required to choose from two natural language statements of similar wording the one that makes sense and the one does not. The second subtask additionally asks a system to select the key reason from three options why a given statement does not make sense. In the third subtask, a participating system needs to generate the reason. We finally attracted 39 teams participating at least one of the three subtasks. For Subtask A and Subtask B, the performances of top-ranked systems are close to that of humans. However, for Subtask C, there is still a relatively large gap between systems and human performance. The dataset used in our task can be found at https://github.com/wangcunxiang/SemEval2020- Task4-Commonsense-Validation-and-Explanation; The leaderboard can be found at https://competitions.codalab.org/competitions/21080#results.	[ { "created": "Wed, 1 Jul 2020 04:41:05 GMT", "version": "v1" }, { "created": "Mon, 3 Aug 2020 15:13:40 GMT", "version": "v2" } ]	2020-08-04	[ [ "Wang", "Cunxiang", "" ], [ "Liang", "Shuailong", "" ], [ "Jin", "Yili", "" ], [ "Wang", "Yilong", "" ], [ "Zhu", "Xiaodan", "" ], [ "Zhang", "Yue", "" ] ]	In this paper, we present SemEval-2020 Task 4, Commonsense Validation and Explanation (ComVE), which includes three subtasks, aiming to evaluate whether a system can distinguish a natural language statement that makes sense to humans from one that does not, and provide the reasons. Specifically, in our first subtask, the participating systems are required to choose from two natural language statements of similar wording the one that makes sense and the one does not. The second subtask additionally asks a system to select the key reason from three options why a given statement does not make sense. In the third subtask, a participating system needs to generate the reason. We finally attracted 39 teams participating at least one of the three subtasks. For Subtask A and Subtask B, the performances of top-ranked systems are close to that of humans. However, for Subtask C, there is still a relatively large gap between systems and human performance. The dataset used in our task can be found at https://github.com/wangcunxiang/SemEval2020- Task4-Commonsense-Validation-and-Explanation; The leaderboard can be found at https://competitions.codalab.org/competitions/21080#results.
2304.13518	Yuqi Han	Yuqi Han and Tao Yu and Xiaohang Yu and Yuwang Wang and Qionghai Dai	Super-NeRF: View-consistent Detail Generation for NeRF super-resolution	null	null	null	null	cs.CV	http://creativecommons.org/licenses/by/4.0/	The neural radiance field (NeRF) achieved remarkable success in modeling 3D scenes and synthesizing high-fidelity novel views. However, existing NeRF-based methods focus more on the make full use of the image resolution to generate novel views, but less considering the generation of details under the limited input resolution. In analogy to the extensive usage of image super-resolution, NeRF super-resolution is an effective way to generate the high-resolution implicit representation of 3D scenes and holds great potential applications. Up to now, such an important topic is still under-explored. In this paper, we propose a NeRF super-resolution method, named Super-NeRF, to generate high-resolution NeRF from only low-resolution inputs. Given multi-view low-resolution images, Super-NeRF constructs a consistency-controlling super-resolution module to generate view-consistent high-resolution details for NeRF. Specifically, an optimizable latent code is introduced for each low-resolution input image to control the 2D super-resolution images to converge to the view-consistent output. The latent codes of each low-resolution image are optimized synergistically with the target Super-NeRF representation to fully utilize the view consistency constraint inherent in NeRF construction. We verify the effectiveness of Super-NeRF on synthetic, real-world, and AI-generated NeRF datasets. Super-NeRF achieves state-of-the-art NeRF super-resolution performance on high-resolution detail generation and cross-view consistency.	[ { "created": "Wed, 26 Apr 2023 12:54:40 GMT", "version": "v1" } ]	2023-04-27	[ [ "Han", "Yuqi", "" ], [ "Yu", "Tao", "" ], [ "Yu", "Xiaohang", "" ], [ "Wang", "Yuwang", "" ], [ "Dai", "Qionghai", "" ] ]	The neural radiance field (NeRF) achieved remarkable success in modeling 3D scenes and synthesizing high-fidelity novel views. However, existing NeRF-based methods focus more on the make full use of the image resolution to generate novel views, but less considering the generation of details under the limited input resolution. In analogy to the extensive usage of image super-resolution, NeRF super-resolution is an effective way to generate the high-resolution implicit representation of 3D scenes and holds great potential applications. Up to now, such an important topic is still under-explored. In this paper, we propose a NeRF super-resolution method, named Super-NeRF, to generate high-resolution NeRF from only low-resolution inputs. Given multi-view low-resolution images, Super-NeRF constructs a consistency-controlling super-resolution module to generate view-consistent high-resolution details for NeRF. Specifically, an optimizable latent code is introduced for each low-resolution input image to control the 2D super-resolution images to converge to the view-consistent output. The latent codes of each low-resolution image are optimized synergistically with the target Super-NeRF representation to fully utilize the view consistency constraint inherent in NeRF construction. We verify the effectiveness of Super-NeRF on synthetic, real-world, and AI-generated NeRF datasets. Super-NeRF achieves state-of-the-art NeRF super-resolution performance on high-resolution detail generation and cross-view consistency.
1809.05606	Yimin Yang	Yimin Yang, Q.M.Jonathan Wu, Xiexing Feng, Thangarajah Akilan	Non-iterative recomputation of dense layers for performance improvement of DCNN	11	null	null	null	cs.LG stat.ML	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	An iterative method of learning has become a paradigm for training deep convolutional neural networks (DCNN). However, utilizing a non-iterative learning strategy can accelerate the training process of the DCNN and surprisingly such approach has been rarely explored by the deep learning (DL) community. It motivates this paper to introduce a non-iterative learning strategy that eliminates the backpropagation (BP) at the top dense or fully connected (FC) layers of DCNN, resulting in, lower training time and higher performance. The proposed method exploits the Moore-Penrose Inverse to pull back the current residual error to each FC layer, generating well-generalized features. Then using the recomputed features, i.e., the new generalized features the weights of each FC layer is computed according to the Moore-Penrose Inverse. We evaluate the proposed approach on six widely accepted object recognition benchmark datasets: Scene-15, CIFAR-10, CIFAR-100, SUN-397, Places365, and ImageNet. The experimental results show that the proposed method obtains significant improvements over 30 state-of-the-art methods. Interestingly, it also indicates that any DCNN with the proposed method can provide better performance than the same network with its original training based on BP.	[ { "created": "Fri, 14 Sep 2018 22:24:52 GMT", "version": "v1" } ]	2018-09-18	[ [ "Yang", "Yimin", "" ], [ "Wu", "Q. M. Jonathan", "" ], [ "Feng", "Xiexing", "" ], [ "Akilan", "Thangarajah", "" ] ]	An iterative method of learning has become a paradigm for training deep convolutional neural networks (DCNN). However, utilizing a non-iterative learning strategy can accelerate the training process of the DCNN and surprisingly such approach has been rarely explored by the deep learning (DL) community. It motivates this paper to introduce a non-iterative learning strategy that eliminates the backpropagation (BP) at the top dense or fully connected (FC) layers of DCNN, resulting in, lower training time and higher performance. The proposed method exploits the Moore-Penrose Inverse to pull back the current residual error to each FC layer, generating well-generalized features. Then using the recomputed features, i.e., the new generalized features the weights of each FC layer is computed according to the Moore-Penrose Inverse. We evaluate the proposed approach on six widely accepted object recognition benchmark datasets: Scene-15, CIFAR-10, CIFAR-100, SUN-397, Places365, and ImageNet. The experimental results show that the proposed method obtains significant improvements over 30 state-of-the-art methods. Interestingly, it also indicates that any DCNN with the proposed method can provide better performance than the same network with its original training based on BP.
1309.0871	EPTCS	Oded Maler (CNRS-VERIMAG, University of Grenoble), \'Ad\'am M. Hal\'asz (Department of Methematics, West Virginia University), Olivier Lebeltel (CNRS-VERIMAG, University of Grenoble), Ouri Maler (Grenoble)	Exploring the Dynamics of Mass Action Systems	In Proceedings HSB 2013, arXiv:1308.5724	EPTCS 125, 2013, pp. 84-91	10.4204/EPTCS.125.6	null	cs.CE cs.CY	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	We present the Populus toolkit for exploring the dynamics of mass action systems under different assumptions.	[ { "created": "Tue, 3 Sep 2013 23:41:36 GMT", "version": "v1" } ]	2013-09-05	[ [ "Maler", "Oded", "", "CNRS-VERIMAG, University of Grenoble" ], [ "Halász", "Ádám M.", "", "Department of Methematics, West Virginia University" ], [ "Lebeltel", "Olivier", "", "CNRS-VERIMAG, University of Grenoble" ], [ "Maler", "Ouri", "", "Grenoble" ] ]	We present the Populus toolkit for exploring the dynamics of mass action systems under different assumptions.
2403.05193	Martina Benini	Martina Benini, Silvia Gallucci, Marta Bonato, Marta Parazzini, Gabriella Tognola	Evaluation of Road User Radio-Frequency Exposure Levels in an Urban Environment from Vehicular Antennas and the Infrastructure in ITS-G5 5.9 GHz Communication	null	null	null	null	cs.NI	http://creativecommons.org/licenses/by/4.0/	This study aims to investigate the variability of exposure levels among road users generated in a realistic urban scenario by Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies operating at 5.9 GHz. The exposure levels were evaluated in terms of whole-body Specific Absorption Rate (wbSAR) [W/kg] in three different human models, ranging from children to adults. We calculated the electromagnetic field exposure level generated by V2V and V2I using raytracing and we assessed wbSAR resulting in urban exposure scenarios with an increasing number of transmitting antennas. Whole-body SAR was generally very low, on the order of 10^-4 W/kg. The maximum wbSAR, of 4.9x10^-4 W/kg, was obtained in the worst-case exposure condition comprising more than one transmitting vehicle and was found in the adult model for a distance within 10 m from the transmitting cars. We found that the height of the human model highly impacted the exposure level. Namely, the child (which is the shortest human model) was generally much less exposed than adults. All the wbSAR values found by varying the number of transmitting antennas, the distance of the road user from the antennas, and the type of human model (adult vs. child) were very well below the limits set by the ICNIRP and IEEE guidelines of 0.08 W/kg for human exposure in the 100 kHz - 300 GHz range.	[ { "created": "Fri, 8 Mar 2024 10:14:40 GMT", "version": "v1" } ]	2024-03-11	[ [ "Benini", "Martina", "" ], [ "Gallucci", "Silvia", "" ], [ "Bonato", "Marta", "" ], [ "Parazzini", "Marta", "" ], [ "Tognola", "Gabriella", "" ] ]	This study aims to investigate the variability of exposure levels among road users generated in a realistic urban scenario by Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies operating at 5.9 GHz. The exposure levels were evaluated in terms of whole-body Specific Absorption Rate (wbSAR) [W/kg] in three different human models, ranging from children to adults. We calculated the electromagnetic field exposure level generated by V2V and V2I using raytracing and we assessed wbSAR resulting in urban exposure scenarios with an increasing number of transmitting antennas. Whole-body SAR was generally very low, on the order of 10^-4 W/kg. The maximum wbSAR, of 4.9x10^-4 W/kg, was obtained in the worst-case exposure condition comprising more than one transmitting vehicle and was found in the adult model for a distance within 10 m from the transmitting cars. We found that the height of the human model highly impacted the exposure level. Namely, the child (which is the shortest human model) was generally much less exposed than adults. All the wbSAR values found by varying the number of transmitting antennas, the distance of the road user from the antennas, and the type of human model (adult vs. child) were very well below the limits set by the ICNIRP and IEEE guidelines of 0.08 W/kg for human exposure in the 100 kHz - 300 GHz range.
2408.03945	Kristina Schaaff	Kristina Schaaff and Marc-Andr\'e Heidelmann	Impacts of Anthropomorphizing Large Language Models in Learning Environments	Presented at Affective Computing Pre-Conference at ISRE 2024	null	null	null	cs.CL cs.AI cs.CY cs.HC	http://creativecommons.org/licenses/by/4.0/	Large Language Models (LLMs) are increasingly being used in learning environments to support teaching-be it as learning companions or as tutors. With our contribution, we aim to discuss the implications of the anthropomorphization of LLMs in learning environments on educational theory to build a foundation for more effective learning outcomes and understand their emotional impact on learners. According to the media equation, people tend to respond to media in the same way as they would respond to another person. A study conducted by the Georgia Institute of Technology showed that chatbots can be successfully implemented in learning environments. In this study, learners in selected online courses were unable to distinguish the chatbot from a "real" teacher. As LLM-based chatbots such as OpenAI's GPT series are increasingly used in educational tools, it is important to understand how the attribution processes to LLM-based chatbots in terms of anthropomorphization affect learners' emotions.	[ { "created": "Mon, 22 Jul 2024 06:28:54 GMT", "version": "v1" } ]	2024-08-09	[ [ "Schaaff", "Kristina", "" ], [ "Heidelmann", "Marc-André", "" ] ]	Large Language Models (LLMs) are increasingly being used in learning environments to support teaching-be it as learning companions or as tutors. With our contribution, we aim to discuss the implications of the anthropomorphization of LLMs in learning environments on educational theory to build a foundation for more effective learning outcomes and understand their emotional impact on learners. According to the media equation, people tend to respond to media in the same way as they would respond to another person. A study conducted by the Georgia Institute of Technology showed that chatbots can be successfully implemented in learning environments. In this study, learners in selected online courses were unable to distinguish the chatbot from a "real" teacher. As LLM-based chatbots such as OpenAI's GPT series are increasingly used in educational tools, it is important to understand how the attribution processes to LLM-based chatbots in terms of anthropomorphization affect learners' emotions.
2310.04197	Angel Casanova	\'Angel Casanova Bienzobas, Alfonso S\'anchez-Maci\'an	Threat Trekker: An Approach to Cyber Threat Hunting	I am disseminating this outcome to all of you, despite the fact that the results may appear somewhat idealistic, given that certain datasets utilized for the training of the machine learning model comprise simulated data	null	null	null	cs.CR	http://creativecommons.org/licenses/by-nc-nd/4.0/	Threat hunting is a proactive methodology for exploring, detecting and mitigating cyberattacks within complex environments. As opposed to conventional detection systems, threat hunting strategies assume adversaries have infiltrated the system; as a result they proactively search out any unusual patterns or activities which might indicate intrusion attempts. Historically, this endeavour has been pursued using three investigation methodologies: (1) Hypothesis-Driven Investigations; (2) Indicator of Compromise (IOC); and (3) High-level machine learning analysis-based approaches. Therefore, this paper introduces a novel machine learning paradigm known as Threat Trekker. This proposal utilizes connectors to feed data directly into an event streaming channel for processing by the algorithm and provide feedback back into its host network. Conclusions drawn from these experiments clearly establish the efficacy of employing machine learning for classifying more subtle attacks.	[ { "created": "Fri, 6 Oct 2023 12:29:41 GMT", "version": "v1" } ]	2023-10-09	[ [ "Bienzobas", "Ángel Casanova", "" ], [ "Sánchez-Macián", "Alfonso", "" ] ]	Threat hunting is a proactive methodology for exploring, detecting and mitigating cyberattacks within complex environments. As opposed to conventional detection systems, threat hunting strategies assume adversaries have infiltrated the system; as a result they proactively search out any unusual patterns or activities which might indicate intrusion attempts. Historically, this endeavour has been pursued using three investigation methodologies: (1) Hypothesis-Driven Investigations; (2) Indicator of Compromise (IOC); and (3) High-level machine learning analysis-based approaches. Therefore, this paper introduces a novel machine learning paradigm known as Threat Trekker. This proposal utilizes connectors to feed data directly into an event streaming channel for processing by the algorithm and provide feedback back into its host network. Conclusions drawn from these experiments clearly establish the efficacy of employing machine learning for classifying more subtle attacks.
2205.06255	Qianqian Wang	Qianqian Wang, Zhengqi Li, David Salesin, Noah Snavely, Brian Curless, Janne Kontkanen	3D Moments from Near-Duplicate Photos	CVPR 2022	null	null	null	cs.CV	http://creativecommons.org/licenses/by/4.0/	We introduce 3D Moments, a new computational photography effect. As input we take a pair of near-duplicate photos, i.e., photos of moving subjects from similar viewpoints, common in people's photo collections. As output, we produce a video that smoothly interpolates the scene motion from the first photo to the second, while also producing camera motion with parallax that gives a heightened sense of 3D. To achieve this effect, we represent the scene as a pair of feature-based layered depth images augmented with scene flow. This representation enables motion interpolation along with independent control of the camera viewpoint. Our system produces photorealistic space-time videos with motion parallax and scene dynamics, while plausibly recovering regions occluded in the original views. We conduct extensive experiments demonstrating superior performance over baselines on public datasets and in-the-wild photos. Project page: https://3d-moments.github.io/	[ { "created": "Thu, 12 May 2022 17:56:18 GMT", "version": "v1" } ]	2022-05-13	[ [ "Wang", "Qianqian", "" ], [ "Li", "Zhengqi", "" ], [ "Salesin", "David", "" ], [ "Snavely", "Noah", "" ], [ "Curless", "Brian", "" ], [ "Kontkanen", "Janne", "" ] ]	We introduce 3D Moments, a new computational photography effect. As input we take a pair of near-duplicate photos, i.e., photos of moving subjects from similar viewpoints, common in people's photo collections. As output, we produce a video that smoothly interpolates the scene motion from the first photo to the second, while also producing camera motion with parallax that gives a heightened sense of 3D. To achieve this effect, we represent the scene as a pair of feature-based layered depth images augmented with scene flow. This representation enables motion interpolation along with independent control of the camera viewpoint. Our system produces photorealistic space-time videos with motion parallax and scene dynamics, while plausibly recovering regions occluded in the original views. We conduct extensive experiments demonstrating superior performance over baselines on public datasets and in-the-wild photos. Project page: https://3d-moments.github.io/
2407.15373	Dizhi Ma	Dizhi Ma, Xiyun Hu, Jingyu Shi, Mayank Patel, Rahul Jain, Ziyi Liu, Zhengzhe Zhu and Karthik Ramani	avaTTAR: Table Tennis Stroke Training with On-body and Detached Visualization in Augmented Reality	null	null	null	null	cs.HC	http://creativecommons.org/licenses/by/4.0/	Table tennis stroke training is a critical aspect of player development. We designed a new augmented reality (AR) system, avaTTAR, for table tennis stroke training. The system provides both "on-body" (first-person view) and "detached" (third-person view) visual cues, enabling users to visualize target strokes and correct their attempts effectively with this dual perspectives setup. By employing a combination of pose estimation algorithms and IMU sensors, avaTTAR captures and reconstructs the 3D body pose and paddle orientation of users during practice, allowing real-time comparison with expert strokes. Through a user study, we affirm avaTTAR's capacity to amplify player experience and training results.	[ { "created": "Mon, 22 Jul 2024 04:47:16 GMT", "version": "v1" }, { "created": "Fri, 26 Jul 2024 15:13:46 GMT", "version": "v2" } ]	2024-07-29	[ [ "Ma", "Dizhi", "" ], [ "Hu", "Xiyun", "" ], [ "Shi", "Jingyu", "" ], [ "Patel", "Mayank", "" ], [ "Jain", "Rahul", "" ], [ "Liu", "Ziyi", "" ], [ "Zhu", "Zhengzhe", "" ], [ "Ramani", "Karthik", "" ] ]	Table tennis stroke training is a critical aspect of player development. We designed a new augmented reality (AR) system, avaTTAR, for table tennis stroke training. The system provides both "on-body" (first-person view) and "detached" (third-person view) visual cues, enabling users to visualize target strokes and correct their attempts effectively with this dual perspectives setup. By employing a combination of pose estimation algorithms and IMU sensors, avaTTAR captures and reconstructs the 3D body pose and paddle orientation of users during practice, allowing real-time comparison with expert strokes. Through a user study, we affirm avaTTAR's capacity to amplify player experience and training results.
2005.13102	Bangalore Ravi Kiran	Leonardo Gigli, B Ravi Kiran, Thomas Paul, Andres Serna, Nagarjuna Vemuri, Beatriz Marcotegui, Santiago Velasco-Forero	Road Segmentation on low resolution Lidar point clouds for autonomous vehicles	ISPRS 2020	null	null	null	cs.CV stat.ML	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Point cloud datasets for perception tasks in the context of autonomous driving often rely on high resolution 64-layer Light Detection and Ranging (LIDAR) scanners. They are expensive to deploy on real-world autonomous driving sensor architectures which usually employ 16/32 layer LIDARs. We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. In our experiments the low resolution 16/32 layer LIDAR point clouds are simulated by subsampling the original 64 layer data, for subsequent transformation in to a feature map in the Bird-Eye-View (BEV) and SphericalView (SV) representations of the point cloud. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures. We demonstrate that this local normal feature in conjunction with classical features not only improves performance for binary road segmentation on full resolution point clouds, but it also reduces the negative impact on the accuracy when subsampling dense point clouds as compared to the usage of classical features alone. We assess our method with several experiments on two datasets: KITTI Road-segmentation benchmark and the recently released Semantic KITTI dataset.	[ { "created": "Wed, 27 May 2020 00:38:39 GMT", "version": "v1" } ]	2020-05-28	[ [ "Gigli", "Leonardo", "" ], [ "Kiran", "B Ravi", "" ], [ "Paul", "Thomas", "" ], [ "Serna", "Andres", "" ], [ "Vemuri", "Nagarjuna", "" ], [ "Marcotegui", "Beatriz", "" ], [ "Velasco-Forero", "Santiago", "" ] ]	Point cloud datasets for perception tasks in the context of autonomous driving often rely on high resolution 64-layer Light Detection and Ranging (LIDAR) scanners. They are expensive to deploy on real-world autonomous driving sensor architectures which usually employ 16/32 layer LIDARs. We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. In our experiments the low resolution 16/32 layer LIDAR point clouds are simulated by subsampling the original 64 layer data, for subsequent transformation in to a feature map in the Bird-Eye-View (BEV) and SphericalView (SV) representations of the point cloud. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures. We demonstrate that this local normal feature in conjunction with classical features not only improves performance for binary road segmentation on full resolution point clouds, but it also reduces the negative impact on the accuracy when subsampling dense point clouds as compared to the usage of classical features alone. We assess our method with several experiments on two datasets: KITTI Road-segmentation benchmark and the recently released Semantic KITTI dataset.
2105.05026	Chongxuan Li	Guoqiang Wu, Chongxuan Li, Kun Xu, Jun Zhu	Rethinking and Reweighting the Univariate Losses for Multi-Label Ranking: Consistency and Generalization	null	null	null	null	cs.LG stat.ML	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	(Partial) ranking loss is a commonly used evaluation measure for multi-label classification, which is usually optimized with convex surrogates for computational efficiency. Prior theoretical work on multi-label ranking mainly focuses on (Fisher) consistency analyses. However, there is a gap between existing theory and practice -- some pairwise losses can lead to promising performance but lack consistency, while some univariate losses are consistent but usually have no clear superiority in practice. In this paper, we attempt to fill this gap through a systematic study from two complementary perspectives of consistency and generalization error bounds of learning algorithms. Our results show that learning algorithms with the consistent univariate loss have an error bound of $O(c)$ ($c$ is the number of labels), while algorithms with the inconsistent pairwise loss depend on $O(\sqrt{c})$ as shown in prior work. This explains that the latter can achieve better performance than the former in practice. Moreover, we present an inconsistent reweighted univariate loss-based learning algorithm that enjoys an error bound of $O(\sqrt{c})$ for promising performance as well as the computational efficiency of univariate losses. Finally, experimental results validate our theoretical analyses.	[ { "created": "Mon, 10 May 2021 09:23:27 GMT", "version": "v1" } ]	2021-05-12	[ [ "Wu", "Guoqiang", "" ], [ "Li", "Chongxuan", "" ], [ "Xu", "Kun", "" ], [ "Zhu", "Jun", "" ] ]	(Partial) ranking loss is a commonly used evaluation measure for multi-label classification, which is usually optimized with convex surrogates for computational efficiency. Prior theoretical work on multi-label ranking mainly focuses on (Fisher) consistency analyses. However, there is a gap between existing theory and practice -- some pairwise losses can lead to promising performance but lack consistency, while some univariate losses are consistent but usually have no clear superiority in practice. In this paper, we attempt to fill this gap through a systematic study from two complementary perspectives of consistency and generalization error bounds of learning algorithms. Our results show that learning algorithms with the consistent univariate loss have an error bound of $O(c)$ ($c$ is the number of labels), while algorithms with the inconsistent pairwise loss depend on $O(\sqrt{c})$ as shown in prior work. This explains that the latter can achieve better performance than the former in practice. Moreover, we present an inconsistent reweighted univariate loss-based learning algorithm that enjoys an error bound of $O(\sqrt{c})$ for promising performance as well as the computational efficiency of univariate losses. Finally, experimental results validate our theoretical analyses.
2403.14623	Zhicong Tang	Zhicong Tang, Tiankai Hang, Shuyang Gu, Dong Chen, Baining Guo	Simplified Diffusion Schr\"odinger Bridge	null	null	null	null	cs.LG cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	This paper introduces a novel theoretical simplification of the Diffusion Schr\"odinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance. By employing SGMs as an initial solution for DSB, our approach capitalizes on the strengths of both frameworks, ensuring a more efficient training process and improving the performance of SGM. We also propose a reparameterization technique that, despite theoretical approximations, practically improves the network's fitting capabilities. Our extensive experimental evaluations confirm the effectiveness of the simplified DSB, demonstrating its significant improvements. We believe the contributions of this work pave the way for advanced generative modeling.	[ { "created": "Thu, 21 Mar 2024 17:59:41 GMT", "version": "v1" }, { "created": "Wed, 27 Mar 2024 16:49:35 GMT", "version": "v2" }, { "created": "Mon, 27 May 2024 04:44:22 GMT", "version": "v3" }, { "created": "Tue, 13 Aug 2024 04:34:58 GMT", "version": "v4" } ]	2024-08-14	[ [ "Tang", "Zhicong", "" ], [ "Hang", "Tiankai", "" ], [ "Gu", "Shuyang", "" ], [ "Chen", "Dong", "" ], [ "Guo", "Baining", "" ] ]	This paper introduces a novel theoretical simplification of the Diffusion Schr\"odinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance. By employing SGMs as an initial solution for DSB, our approach capitalizes on the strengths of both frameworks, ensuring a more efficient training process and improving the performance of SGM. We also propose a reparameterization technique that, despite theoretical approximations, practically improves the network's fitting capabilities. Our extensive experimental evaluations confirm the effectiveness of the simplified DSB, demonstrating its significant improvements. We believe the contributions of this work pave the way for advanced generative modeling.
1511.04376	Martin Reisslein	Akhilesh Thyagaturu, Anu Mercian, Michael P. McGarry, Martin Reisslein, Wolfgang Kellerer	Software Defined Optical Networks (SDONs): A Comprehensive Survey	null	IEEE Communications Surveys & Tutorials, vol. 18, no. 4, pp. 2738-2786, 4th Qu. 2016	10.1109/COMST.2016.2586999	null	cs.NI	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	The emerging Software Defined Networking (SDN) paradigm separates the data plane from the control plane and centralizes network control in an SDN controller. Applications interact with controllers to implement network services, such as network transport with Quality of Service (QoS). SDN facilitates the virtualization of network functions so that multiple virtual networks can operate over a given installed physical network infrastructure. Due to the specific characteristics of optical (photonic) communication components and the high optical transmission capacities, SDN based optical networking poses particular challenges, but holds also great potential. In this article, we comprehensively survey studies that examine the SDN paradigm in optical networks; in brief, we survey the area of Software Defined Optical Networks (SDONs). We mainly organize the SDON studies into studies focused on the infrastructure layer, the control layer, and the application layer. Moreover, we cover SDON studies focused on network virtualization, as well as SDON studies focused on the orchestration of multilayer and multidomain networking. Based on the survey, we identify open challenges for SDONs and outline future directions.	[ { "created": "Fri, 13 Nov 2015 17:31:10 GMT", "version": "v1" }, { "created": "Thu, 26 May 2016 10:19:58 GMT", "version": "v2" }, { "created": "Sun, 17 Jul 2016 07:46:47 GMT", "version": "v3" } ]	2016-11-29	[ [ "Thyagaturu", "Akhilesh", "" ], [ "Mercian", "Anu", "" ], [ "McGarry", "Michael P.", "" ], [ "Reisslein", "Martin", "" ], [ "Kellerer", "Wolfgang", "" ] ]	The emerging Software Defined Networking (SDN) paradigm separates the data plane from the control plane and centralizes network control in an SDN controller. Applications interact with controllers to implement network services, such as network transport with Quality of Service (QoS). SDN facilitates the virtualization of network functions so that multiple virtual networks can operate over a given installed physical network infrastructure. Due to the specific characteristics of optical (photonic) communication components and the high optical transmission capacities, SDN based optical networking poses particular challenges, but holds also great potential. In this article, we comprehensively survey studies that examine the SDN paradigm in optical networks; in brief, we survey the area of Software Defined Optical Networks (SDONs). We mainly organize the SDON studies into studies focused on the infrastructure layer, the control layer, and the application layer. Moreover, we cover SDON studies focused on network virtualization, as well as SDON studies focused on the orchestration of multilayer and multidomain networking. Based on the survey, we identify open challenges for SDONs and outline future directions.
2306.02346	Shuo Ye	Shuo Ye and Yufeng Shi and Ruxin Wang and Yu Wang and Jiamiao Xu and Chuanwu Yang and Xinge You	CDLT: A Dataset with Concept Drift and Long-Tailed Distribution for Fine-Grained Visual Categorization	null	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Data is the foundation for the development of computer vision, and the establishment of datasets plays an important role in advancing the techniques of fine-grained visual categorization~(FGVC). In the existing FGVC datasets used in computer vision, it is generally assumed that each collected instance has fixed characteristics and the distribution of different categories is relatively balanced. In contrast, the real world scenario reveals the fact that the characteristics of instances tend to vary with time and exhibit a long-tailed distribution. Hence, the collected datasets may mislead the optimization of the fine-grained classifiers, resulting in unpleasant performance in real applications. Starting from the real-world conditions and to promote the practical progress of fine-grained visual categorization, we present a Concept Drift and Long-Tailed Distribution dataset. Specifically, the dataset is collected by gathering 11195 images of 250 instances in different species for 47 consecutive months in their natural contexts. The collection process involves dozens of crowd workers for photographing and domain experts for labelling. Extensive baseline experiments using the state-of-the-art fine-grained classification models demonstrate the issues of concept drift and long-tailed distribution existed in the dataset, which require the attention of future researches.	[ { "created": "Sun, 4 Jun 2023 12:42:45 GMT", "version": "v1" } ]	2023-06-06	[ [ "Ye", "Shuo", "" ], [ "Shi", "Yufeng", "" ], [ "Wang", "Ruxin", "" ], [ "Wang", "Yu", "" ], [ "Xu", "Jiamiao", "" ], [ "Yang", "Chuanwu", "" ], [ "You", "Xinge", "" ] ]	Data is the foundation for the development of computer vision, and the establishment of datasets plays an important role in advancing the techniques of fine-grained visual categorization~(FGVC). In the existing FGVC datasets used in computer vision, it is generally assumed that each collected instance has fixed characteristics and the distribution of different categories is relatively balanced. In contrast, the real world scenario reveals the fact that the characteristics of instances tend to vary with time and exhibit a long-tailed distribution. Hence, the collected datasets may mislead the optimization of the fine-grained classifiers, resulting in unpleasant performance in real applications. Starting from the real-world conditions and to promote the practical progress of fine-grained visual categorization, we present a Concept Drift and Long-Tailed Distribution dataset. Specifically, the dataset is collected by gathering 11195 images of 250 instances in different species for 47 consecutive months in their natural contexts. The collection process involves dozens of crowd workers for photographing and domain experts for labelling. Extensive baseline experiments using the state-of-the-art fine-grained classification models demonstrate the issues of concept drift and long-tailed distribution existed in the dataset, which require the attention of future researches.
2303.06545	Hao Zhou	Hao Zhou, Chongyang Zhang, Yanjun Chen, Chuanping Hu	Towards Diverse Temporal Grounding under Single Positive Labels	The source codes are available at https://github.com/zhouhaocv/DTG-SPL	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Temporal grounding aims to retrieve moments of the described event within an untrimmed video by a language query. Typically, existing methods assume annotations are precise and unique, yet one query may describe multiple moments in many cases. Hence, simply taking it as a one-vs-one mapping task and striving to match single-label annotations will inevitably introduce false negatives during optimization. In this study, we reformulate this task as a one-vs-many optimization problem under the condition of single positive labels. The unlabeled moments are considered unobserved rather than negative, and we explore mining potential positive moments to assist in multiple moment retrieval. In this setting, we propose a novel Diverse Temporal Grounding framework, termed DTG-SPL, which mainly consists of a positive moment estimation (PME) module and a diverse moment regression (DMR) module. PME leverages semantic reconstruction information and an expected positive regularization to uncover potential positive moments in an online fashion. Under the supervision of these pseudo positives, DMR is able to localize diverse moments in parallel that meet different users. The entire framework allows for end-to-end optimization as well as fast inference. Extensive experiments on Charades-STA and ActivityNet Captions show that our method achieves superior performance in terms of both single-label and multi-label metrics.	[ { "created": "Sun, 12 Mar 2023 02:54:18 GMT", "version": "v1" } ]	2023-03-14	[ [ "Zhou", "Hao", "" ], [ "Zhang", "Chongyang", "" ], [ "Chen", "Yanjun", "" ], [ "Hu", "Chuanping", "" ] ]	Temporal grounding aims to retrieve moments of the described event within an untrimmed video by a language query. Typically, existing methods assume annotations are precise and unique, yet one query may describe multiple moments in many cases. Hence, simply taking it as a one-vs-one mapping task and striving to match single-label annotations will inevitably introduce false negatives during optimization. In this study, we reformulate this task as a one-vs-many optimization problem under the condition of single positive labels. The unlabeled moments are considered unobserved rather than negative, and we explore mining potential positive moments to assist in multiple moment retrieval. In this setting, we propose a novel Diverse Temporal Grounding framework, termed DTG-SPL, which mainly consists of a positive moment estimation (PME) module and a diverse moment regression (DMR) module. PME leverages semantic reconstruction information and an expected positive regularization to uncover potential positive moments in an online fashion. Under the supervision of these pseudo positives, DMR is able to localize diverse moments in parallel that meet different users. The entire framework allows for end-to-end optimization as well as fast inference. Extensive experiments on Charades-STA and ActivityNet Captions show that our method achieves superior performance in terms of both single-label and multi-label metrics.
2110.15919	Pranay Bhardwaj	Vinay U. Pai, Pranay Bhardwaj, and S. M. Zafaruddin	Performance Analysis of Dual-Hop THz Wireless Transmission for Backhaul Applications	This paper has been accepted for presentation in 2021 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS), Hyderabad, India	null	null	null	cs.IT eess.SP math.IT	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	THz transmissions suffer from pointing errors due to antenna misalignment and incur higher path loss from the molecular absorption in addition to the channel fading. In this paper, we employ an amplify-and-forward (AF) dual-hop relaying to mitigate the effect of pointing errors and extend the range of the THz wireless system for backhaul connectivity. We provide statistical analysis on the performance of the considered system by deriving analytical expressions for the outage probability, average bit-error-rate (BER), average signal-to-noise ratio (SNR), and a lower bound on the ergodic capacity over independent and identical (i.i.d) $\alpha$-$\mu$ fading combined with the statistical effect of pointing errors. Using computer simulations, we validate the derived analysis of the relay-assisted system. We also demonstrate the effect of the system parameters on outage probability and average BER with the help of diversity order. We show that data rates up to several \mbox{Gbps} can be achieved using THz transmissions, which is desirable for next-generation wireless systems, especially for backhaul applications.	[ { "created": "Fri, 29 Oct 2021 17:15:38 GMT", "version": "v1" }, { "created": "Sun, 21 Nov 2021 19:24:23 GMT", "version": "v2" } ]	2021-11-23	[ [ "Pai", "Vinay U.", "" ], [ "Bhardwaj", "Pranay", "" ], [ "Zafaruddin", "S. M.", "" ] ]	THz transmissions suffer from pointing errors due to antenna misalignment and incur higher path loss from the molecular absorption in addition to the channel fading. In this paper, we employ an amplify-and-forward (AF) dual-hop relaying to mitigate the effect of pointing errors and extend the range of the THz wireless system for backhaul connectivity. We provide statistical analysis on the performance of the considered system by deriving analytical expressions for the outage probability, average bit-error-rate (BER), average signal-to-noise ratio (SNR), and a lower bound on the ergodic capacity over independent and identical (i.i.d) $\alpha$-$\mu$ fading combined with the statistical effect of pointing errors. Using computer simulations, we validate the derived analysis of the relay-assisted system. We also demonstrate the effect of the system parameters on outage probability and average BER with the help of diversity order. We show that data rates up to several \mbox{Gbps} can be achieved using THz transmissions, which is desirable for next-generation wireless systems, especially for backhaul applications.
2202.02071	Henrique Moniz	Afonso Oliveira, Henrique Moniz, Rodrigo Rodrigues	Alea-BFT: Practical Asynchronous Byzantine Fault Tolerance	null	null	null	null	cs.DC	http://creativecommons.org/licenses/by/4.0/	Traditional Byzantine Fault Tolerance (BFT) state machine replication protocols assume a partial synchrony model, leading to a design where a leader replica drives the protocol and is replaced after a timeout. Recently, we witnessed a surge of asynchronous BFT protocols that use randomization to remove the assumptions of bounds on message delivery times, making them more resilient to adverse network conditions. However, these protocols still fall short of being practical across a broad range of scenarios due to their cubic communication costs, use of expensive primitives, and overall protocol complexity. In this paper, we present Alea-BFT, the first asynchronous BFT protocol to achieve quadratic communication complexity, allowing it to scale to large networks. Alea-BFT brings the key design insight from classical protocols of concentrating part of the work on a single designated replica, and incorporates this principle in a two stage pipelined design, with an efficient broadcast led by the designated replica followed by an inexpensive binary agreement. We evaluated our prototype implementation across 10 sites in 4 continents, and our results show significant scalability gains from the proposed design.	[ { "created": "Fri, 4 Feb 2022 10:53:37 GMT", "version": "v1" } ]	2022-02-07	[ [ "Oliveira", "Afonso", "" ], [ "Moniz", "Henrique", "" ], [ "Rodrigues", "Rodrigo", "" ] ]	Traditional Byzantine Fault Tolerance (BFT) state machine replication protocols assume a partial synchrony model, leading to a design where a leader replica drives the protocol and is replaced after a timeout. Recently, we witnessed a surge of asynchronous BFT protocols that use randomization to remove the assumptions of bounds on message delivery times, making them more resilient to adverse network conditions. However, these protocols still fall short of being practical across a broad range of scenarios due to their cubic communication costs, use of expensive primitives, and overall protocol complexity. In this paper, we present Alea-BFT, the first asynchronous BFT protocol to achieve quadratic communication complexity, allowing it to scale to large networks. Alea-BFT brings the key design insight from classical protocols of concentrating part of the work on a single designated replica, and incorporates this principle in a two stage pipelined design, with an efficient broadcast led by the designated replica followed by an inexpensive binary agreement. We evaluated our prototype implementation across 10 sites in 4 continents, and our results show significant scalability gains from the proposed design.
1405.2066	Jehad Al Dallal Prof.	Jehad Al Dallal	How and When to Flatten Java Classes?	null	null	null	null	cs.SE	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Improving modularity and reusability are two key objectives in object-oriented programming. These objectives are achieved by applying several key concepts, such as data encapsulation and inheritance. A class in an object-oriented system is the basic unit of design. Assessing the quality of an object-oriented class may require flattening the class and representing it as it really is, including all accessible inherited class members. Thus, class flattening helps in exploring the impact of inheritance on improving code quality. This paper explains how to flatten Java classes and discusses the relationship between class flattening and some applications of interest to software practitioners, such as refactoring and indicating external quality attributes.	[ { "created": "Thu, 8 May 2014 19:48:33 GMT", "version": "v1" } ]	2014-05-09	[ [ "Dallal", "Jehad Al", "" ] ]	Improving modularity and reusability are two key objectives in object-oriented programming. These objectives are achieved by applying several key concepts, such as data encapsulation and inheritance. A class in an object-oriented system is the basic unit of design. Assessing the quality of an object-oriented class may require flattening the class and representing it as it really is, including all accessible inherited class members. Thus, class flattening helps in exploring the impact of inheritance on improving code quality. This paper explains how to flatten Java classes and discusses the relationship between class flattening and some applications of interest to software practitioners, such as refactoring and indicating external quality attributes.
1406.3583	Aaron D. Jaggard	Aaron D. Jaggard, Aaron Johnson, Paul Syverson, and Joan Feigenbaum	Representing Network Trust and Using It to Improve Anonymous Communication	24 pages; talk to be presented at HotPETs 2014	null	null	null	cs.CR	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Motivated by the effectiveness of correlation attacks against Tor, the censorship arms race, and observations of malicious relays in Tor, we propose that Tor users capture their trust in network elements using probability distributions over the sets of elements observed by network adversaries. We present a modular system that allows users to efficiently and conveniently create such distributions and use them to improve their security. The major components of this system are (i) an ontology of network-element types that represents the main threats to and vulnerabilities of anonymous communication over Tor, (ii) a formal language that allows users to naturally express trust beliefs about network elements, and (iii) a conversion procedure that takes the ontology, public information about the network, and user beliefs written in the trust language and produce a Bayesian Belief Network that represents the probability distribution in a way that is concise and easily sampleable. We also present preliminary experimental results that show the distribution produced by our system can improve security when employed by users; further improvement is seen when the system is employed by both users and services.	[ { "created": "Fri, 13 Jun 2014 16:23:12 GMT", "version": "v1" } ]	2014-06-16	[ [ "Jaggard", "Aaron D.", "" ], [ "Johnson", "Aaron", "" ], [ "Syverson", "Paul", "" ], [ "Feigenbaum", "Joan", "" ] ]	Motivated by the effectiveness of correlation attacks against Tor, the censorship arms race, and observations of malicious relays in Tor, we propose that Tor users capture their trust in network elements using probability distributions over the sets of elements observed by network adversaries. We present a modular system that allows users to efficiently and conveniently create such distributions and use them to improve their security. The major components of this system are (i) an ontology of network-element types that represents the main threats to and vulnerabilities of anonymous communication over Tor, (ii) a formal language that allows users to naturally express trust beliefs about network elements, and (iii) a conversion procedure that takes the ontology, public information about the network, and user beliefs written in the trust language and produce a Bayesian Belief Network that represents the probability distribution in a way that is concise and easily sampleable. We also present preliminary experimental results that show the distribution produced by our system can improve security when employed by users; further improvement is seen when the system is employed by both users and services.
1702.06235	Will Radford	Andrew Chisholm, Will Radford, Ben Hachey	Learning to generate one-sentence biographies from Wikidata	null	null	null	null	cs.CL	http://creativecommons.org/licenses/by/4.0/	We investigate the generation of one-sentence Wikipedia biographies from facts derived from Wikidata slot-value pairs. We train a recurrent neural network sequence-to-sequence model with attention to select facts and generate textual summaries. Our model incorporates a novel secondary objective that helps ensure it generates sentences that contain the input facts. The model achieves a BLEU score of 41, improving significantly upon the vanilla sequence-to-sequence model and scoring roughly twice that of a simple template baseline. Human preference evaluation suggests the model is nearly as good as the Wikipedia reference. Manual analysis explores content selection, suggesting the model can trade the ability to infer knowledge against the risk of hallucinating incorrect information.	[ { "created": "Tue, 21 Feb 2017 01:30:59 GMT", "version": "v1" } ]	2017-02-22	[ [ "Chisholm", "Andrew", "" ], [ "Radford", "Will", "" ], [ "Hachey", "Ben", "" ] ]	We investigate the generation of one-sentence Wikipedia biographies from facts derived from Wikidata slot-value pairs. We train a recurrent neural network sequence-to-sequence model with attention to select facts and generate textual summaries. Our model incorporates a novel secondary objective that helps ensure it generates sentences that contain the input facts. The model achieves a BLEU score of 41, improving significantly upon the vanilla sequence-to-sequence model and scoring roughly twice that of a simple template baseline. Human preference evaluation suggests the model is nearly as good as the Wikipedia reference. Manual analysis explores content selection, suggesting the model can trade the ability to infer knowledge against the risk of hallucinating incorrect information.
2402.11173	Andrew Lowy	Andrew Lowy, Jonathan Ullman, Stephen J. Wright	How to Make the Gradients Small Privately: Improved Rates for Differentially Private Non-Convex Optimization	null	null	null	null	cs.LG cs.CR math.OC	http://creativecommons.org/licenses/by/4.0/	We provide a simple and flexible framework for designing differentially private algorithms to find approximate stationary points of non-convex loss functions. Our framework is based on using a private approximate risk minimizer to "warm start" another private algorithm for finding stationary points. We use this framework to obtain improved, and sometimes optimal, rates for several classes of non-convex loss functions. First, we obtain improved rates for finding stationary points of smooth non-convex empirical loss functions. Second, we specialize to quasar-convex functions, which generalize star-convex functions and arise in learning dynamical systems and training some neural nets. We achieve the optimal rate for this class. Third, we give an optimal algorithm for finding stationary points of functions satisfying the Kurdyka-Lojasiewicz (KL) condition. For example, over-parameterized neural networks often satisfy this condition. Fourth, we provide new state-of-the-art rates for stationary points of non-convex population loss functions. Fifth, we obtain improved rates for non-convex generalized linear models. A modification of our algorithm achieves nearly the same rates for second-order stationary points of functions with Lipschitz Hessian, improving over the previous state-of-the-art for each of the above problems.	[ { "created": "Sat, 17 Feb 2024 02:42:56 GMT", "version": "v1" } ]	2024-02-20	[ [ "Lowy", "Andrew", "" ], [ "Ullman", "Jonathan", "" ], [ "Wright", "Stephen J.", "" ] ]	We provide a simple and flexible framework for designing differentially private algorithms to find approximate stationary points of non-convex loss functions. Our framework is based on using a private approximate risk minimizer to "warm start" another private algorithm for finding stationary points. We use this framework to obtain improved, and sometimes optimal, rates for several classes of non-convex loss functions. First, we obtain improved rates for finding stationary points of smooth non-convex empirical loss functions. Second, we specialize to quasar-convex functions, which generalize star-convex functions and arise in learning dynamical systems and training some neural nets. We achieve the optimal rate for this class. Third, we give an optimal algorithm for finding stationary points of functions satisfying the Kurdyka-Lojasiewicz (KL) condition. For example, over-parameterized neural networks often satisfy this condition. Fourth, we provide new state-of-the-art rates for stationary points of non-convex population loss functions. Fifth, we obtain improved rates for non-convex generalized linear models. A modification of our algorithm achieves nearly the same rates for second-order stationary points of functions with Lipschitz Hessian, improving over the previous state-of-the-art for each of the above problems.
2401.11946	Jiajun Liu	Jiajun Liu, Lina Tan, Zhili Zhou, Yi Li, Peng Chen	A Dynamic YOLO-Based Sequence-Matching Model for Efficient Coverless Image Steganography	null	null	null	null	cs.CR	http://creativecommons.org/licenses/by/4.0/	Many existing coverless steganography methods establish a mapping relationship between cover images and hidden data. There exists an issue that the number of images stored in the database grows exponentially as the steganographic capacity rises. The need for a high steganographic capacity makes it challenging to build an image database. To improve the image library utilization and anti-attack capability of the steganography system, we present an efficient coverless scheme based on dynamically matched substrings. YOLO is employed for selecting optimal objects, and a mapping dictionary is established between these objects and scrambling factors. With the aid of this dictionary, each image is effectively assigned to a specific scrambling factor, which is used to scramble the receiver's sequence key. To achieve sufficient steganography capability based on a limited image library, all substrings of the scrambled sequences hold the potential to hide data. After completing the secret information matching, the ideal number of stego images will be obtained from the database. According to experimental results, this technology outperforms most previous works on data load, transmission security, and hiding capacity. Under typical geometric attacks, it can recover 79.85\% of secret information on average. Furthermore, only approximately 200 random images are needed to meet a capacity of 19 bits per image.	[ { "created": "Mon, 22 Jan 2024 13:35:27 GMT", "version": "v1" } ]	2024-01-23	[ [ "Liu", "Jiajun", "" ], [ "Tan", "Lina", "" ], [ "Zhou", "Zhili", "" ], [ "Li", "Yi", "" ], [ "Chen", "Peng", "" ] ]	Many existing coverless steganography methods establish a mapping relationship between cover images and hidden data. There exists an issue that the number of images stored in the database grows exponentially as the steganographic capacity rises. The need for a high steganographic capacity makes it challenging to build an image database. To improve the image library utilization and anti-attack capability of the steganography system, we present an efficient coverless scheme based on dynamically matched substrings. YOLO is employed for selecting optimal objects, and a mapping dictionary is established between these objects and scrambling factors. With the aid of this dictionary, each image is effectively assigned to a specific scrambling factor, which is used to scramble the receiver's sequence key. To achieve sufficient steganography capability based on a limited image library, all substrings of the scrambled sequences hold the potential to hide data. After completing the secret information matching, the ideal number of stego images will be obtained from the database. According to experimental results, this technology outperforms most previous works on data load, transmission security, and hiding capacity. Under typical geometric attacks, it can recover 79.85\% of secret information on average. Furthermore, only approximately 200 random images are needed to meet a capacity of 19 bits per image.
2203.10837	Marcos Faundez-Zanuy	K. L\'opez-de-Ipi\~na, Marcos Faundez-Zanuy, Jordi Sol\'e-Casals, Fernando Zelarin, Pilar Calvo	Multi-class versus One-class classifier in spontaneous speech analysis oriented to Alzheimer Disease diagnosis	10 pages, published in International Conference on NONLINEAR SPEECH PROCESSING, NOLISP 2015 jointly organized with the 25th Italian Workshop on Neural Networks, WIRN 2015, held at May 2015, Vietri sul Mare, Salerno, Italy	Recent Advances in Nonlinear Speech Processing. Smart Innovation, Systems and Technologies, vol 48. Springer, Cham 2015	10.1007/978-3-319-28109-4_7	null	cs.SD cs.LG eess.AS	http://creativecommons.org/licenses/by-nc-nd/4.0/	Most of medical developments require the ability to identify samples that are anomalous with respect to a target group or control group, in the sense they could belong to a new, previously unseen class or are not class data. In this case when there are not enough data to train two-class One-class classification appear like an available solution. On the other hand non-linear approaches could give very useful information. The aim of our project is to contribute to earlier diagnosis of AD and better estimates of its severity by using automatic analysis performed through new biomarkers extracted from speech signal. The methods selected in this case are speech biomarkers oriented to Spontaneous Speech and Emotional Response Analysis. In this approach One-class classifiers and two-class classifiers are analyzed. The use of information about outlier and Fractal Dimension features improves the system performance.	[ { "created": "Mon, 21 Mar 2022 09:57:20 GMT", "version": "v1" } ]	2022-03-22	[ [ "López-de-Ipiña", "K.", "" ], [ "Faundez-Zanuy", "Marcos", "" ], [ "Solé-Casals", "Jordi", "" ], [ "Zelarin", "Fernando", "" ], [ "Calvo", "Pilar", "" ] ]	Most of medical developments require the ability to identify samples that are anomalous with respect to a target group or control group, in the sense they could belong to a new, previously unseen class or are not class data. In this case when there are not enough data to train two-class One-class classification appear like an available solution. On the other hand non-linear approaches could give very useful information. The aim of our project is to contribute to earlier diagnosis of AD and better estimates of its severity by using automatic analysis performed through new biomarkers extracted from speech signal. The methods selected in this case are speech biomarkers oriented to Spontaneous Speech and Emotional Response Analysis. In this approach One-class classifiers and two-class classifiers are analyzed. The use of information about outlier and Fractal Dimension features improves the system performance.
2408.06047	Xuanpu Zhang	Xuanpu Zhang and Dan Song and Pengxin Zhan and Qingguo Chen and Zhao Xu and Weihua Luo and Kaifu Zhang and Anan Liu	BooW-VTON: Boosting In-the-Wild Virtual Try-On via Mask-Free Pseudo Data Training	null	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Image-based virtual try-on is an increasingly popular and important task to generate realistic try-on images of specific person. Existing methods always employ an accurate mask to remove the original garment in the source image, thus achieving realistic synthesized images in simple and conventional try-on scenarios based on powerful diffusion model. Therefore, acquiring suitable mask is vital to the try-on performance of these methods. However, obtaining precise inpainting masks, especially for complex wild try-on data containing diverse foreground occlusions and person poses, is not easy as Figure 1-Top shows. This difficulty often results in poor performance in more practical and challenging real-life scenarios, such as the selfie scene shown in Figure 1-Bottom. To this end, we propose a novel training paradigm combined with an efficient data augmentation method to acquire large-scale unpaired training data from wild scenarios, thereby significantly facilitating the try-on performance of our model without the need for additional inpainting masks. Besides, a try-on localization loss is designed to localize a more accurate try-on area to obtain more reasonable try-on results. It is noted that our method only needs the reference cloth image, source pose image and source person image as input, which is more cost-effective and user-friendly compared to existing methods. Extensive qualitative and quantitative experiments have demonstrated superior performance in wild scenarios with such a low-demand input.	[ { "created": "Mon, 12 Aug 2024 10:39:59 GMT", "version": "v1" } ]	2024-08-13	[ [ "Zhang", "Xuanpu", "" ], [ "Song", "Dan", "" ], [ "Zhan", "Pengxin", "" ], [ "Chen", "Qingguo", "" ], [ "Xu", "Zhao", "" ], [ "Luo", "Weihua", "" ], [ "Zhang", "Kaifu", "" ], [ "Liu", "Anan", "" ] ]	Image-based virtual try-on is an increasingly popular and important task to generate realistic try-on images of specific person. Existing methods always employ an accurate mask to remove the original garment in the source image, thus achieving realistic synthesized images in simple and conventional try-on scenarios based on powerful diffusion model. Therefore, acquiring suitable mask is vital to the try-on performance of these methods. However, obtaining precise inpainting masks, especially for complex wild try-on data containing diverse foreground occlusions and person poses, is not easy as Figure 1-Top shows. This difficulty often results in poor performance in more practical and challenging real-life scenarios, such as the selfie scene shown in Figure 1-Bottom. To this end, we propose a novel training paradigm combined with an efficient data augmentation method to acquire large-scale unpaired training data from wild scenarios, thereby significantly facilitating the try-on performance of our model without the need for additional inpainting masks. Besides, a try-on localization loss is designed to localize a more accurate try-on area to obtain more reasonable try-on results. It is noted that our method only needs the reference cloth image, source pose image and source person image as input, which is more cost-effective and user-friendly compared to existing methods. Extensive qualitative and quantitative experiments have demonstrated superior performance in wild scenarios with such a low-demand input.
1604.01186	EPTCS	Denis Firsov, Tarmo Uustalu, Niccol\`o Veltri	Variations on Noetherianness	In Proceedings MSFP 2016, arXiv:1604.00384	EPTCS 207, 2016, pp. 76-88	10.4204/EPTCS.207.4	null	cs.LO cs.PL	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	In constructive mathematics, several nonequivalent notions of finiteness exist. In this paper, we continue the study of Noetherian sets in the dependently typed setting of the Agda programming language. We want to say that a set is Noetherian, if, when we are shown elements from it one after another, we will sooner or later have seen some element twice. This idea can be made precise in a number of ways. We explore the properties and connections of some of the possible encodings. In particular, we show that certain implementations imply decidable equality while others do not, and we construct counterexamples in the latter case. Additionally, we explore the relation between Noetherianness and other notions of finiteness.	[ { "created": "Tue, 5 Apr 2016 09:04:13 GMT", "version": "v1" } ]	2016-04-06	[ [ "Firsov", "Denis", "" ], [ "Uustalu", "Tarmo", "" ], [ "Veltri", "Niccolò", "" ] ]	In constructive mathematics, several nonequivalent notions of finiteness exist. In this paper, we continue the study of Noetherian sets in the dependently typed setting of the Agda programming language. We want to say that a set is Noetherian, if, when we are shown elements from it one after another, we will sooner or later have seen some element twice. This idea can be made precise in a number of ways. We explore the properties and connections of some of the possible encodings. In particular, we show that certain implementations imply decidable equality while others do not, and we construct counterexamples in the latter case. Additionally, we explore the relation between Noetherianness and other notions of finiteness.
2109.06612	Johanna Schmidt	Raphael Sahann, Torsten M\"oller, Johanna Schmidt	Histogram binning revisited with a focus on human perception	Accepted as short paper at VIS 2021. Supplemental material can be found at https://github.com/johanna-schmidt/histogram-binning-revisited	Proceedings of VIS short papers 2021	null	null	cs.HC	http://creativecommons.org/licenses/by-nc-sa/4.0/	This paper presents a quantitative user study to evaluate how well users can visually perceive the underlying data distribution from a histogram representation. We used different sample and bin sizes and four different distributions (uniform, normal, bimodal, and gamma). The study results confirm that, in general, more bins correlate with fewer errors by the viewers. However, upon a certain number of bins, the error rate cannot be improved by adding more bins. By comparing our study results with the outcomes of existing mathematical models for histogram binning (e.g., Sturges' formula, Scott's normal reference rule, the Rice Rule, or Freedman-Diaconis' choice), we can see that most of them overestimate the number of bins necessary to make the distribution visible to a human viewer.	[ { "created": "Tue, 14 Sep 2021 12:08:27 GMT", "version": "v1" } ]	2021-09-15	[ [ "Sahann", "Raphael", "" ], [ "Möller", "Torsten", "" ], [ "Schmidt", "Johanna", "" ] ]	This paper presents a quantitative user study to evaluate how well users can visually perceive the underlying data distribution from a histogram representation. We used different sample and bin sizes and four different distributions (uniform, normal, bimodal, and gamma). The study results confirm that, in general, more bins correlate with fewer errors by the viewers. However, upon a certain number of bins, the error rate cannot be improved by adding more bins. By comparing our study results with the outcomes of existing mathematical models for histogram binning (e.g., Sturges' formula, Scott's normal reference rule, the Rice Rule, or Freedman-Diaconis' choice), we can see that most of them overestimate the number of bins necessary to make the distribution visible to a human viewer.
1802.09657	Dipankar Maity	Dipankar Maity, John S. Baras	Event-Triggered Controller Synthesis for Dynamical Systems with Temporal Logic Constraints	null	null	null	null	cs.RO math.DS	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	In this work, we propose an event-triggered con- trol framework for dynamical systems with temporal logical constraints. Event-triggered control methodologies have proven to be very efficient in reducing sensing, communication and computation costs. When a continuous feedback control is re- placed with an event-triggered strategy, the corresponding state trajectories also differ. In a system with logical constraints, such small deviation in the trajectory might lead to unsatisfiability of the logical constraints. In this work, we develop an approach where we ensure that the event-triggered state trajectory is confined within an tube of the ideal trajectory associated with the continuous state feedback. At the same time, we will ensure satisfiability of the logical constraints as well. Furthermore, we show that the proposed method works for delayed systems as long as the delay is bounded by a certain quantity.	[ { "created": "Tue, 27 Feb 2018 00:17:41 GMT", "version": "v1" } ]	2018-02-28	[ [ "Maity", "Dipankar", "" ], [ "Baras", "John S.", "" ] ]	In this work, we propose an event-triggered con- trol framework for dynamical systems with temporal logical constraints. Event-triggered control methodologies have proven to be very efficient in reducing sensing, communication and computation costs. When a continuous feedback control is re- placed with an event-triggered strategy, the corresponding state trajectories also differ. In a system with logical constraints, such small deviation in the trajectory might lead to unsatisfiability of the logical constraints. In this work, we develop an approach where we ensure that the event-triggered state trajectory is confined within an tube of the ideal trajectory associated with the continuous state feedback. At the same time, we will ensure satisfiability of the logical constraints as well. Furthermore, we show that the proposed method works for delayed systems as long as the delay is bounded by a certain quantity.
1911.03059	Chowdhury Rahman	Afra Anika, Md. Hasibur Rahman, Salekul Islam, Abu Shafin Mohammad Mahdee Jameel and Chowdhury Rafeed Rahman	A Comprehensive Comparison of Machine Learning Based Methods Used in Bengali Question Classification	null	null	null	null	cs.CL	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	QA classification system maps questions asked by humans to an appropriate answer category. A sound question classification (QC) system model is the pre-requisite of a sound QA system. This work demonstrates phases of assembling a QA type classification model. We present a comprehensive comparison (performance and computational complexity) among some machine learning based approaches used in QC for Bengali language.	[ { "created": "Fri, 8 Nov 2019 05:30:33 GMT", "version": "v1" }, { "created": "Tue, 19 Nov 2019 16:37:41 GMT", "version": "v2" } ]	2019-11-20	[ [ "Anika", "Afra", "" ], [ "Rahman", "Md. Hasibur", "" ], [ "Islam", "Salekul", "" ], [ "Jameel", "Abu Shafin Mohammad Mahdee", "" ], [ "Rahman", "Chowdhury Rafeed", "" ] ]	QA classification system maps questions asked by humans to an appropriate answer category. A sound question classification (QC) system model is the pre-requisite of a sound QA system. This work demonstrates phases of assembling a QA type classification model. We present a comprehensive comparison (performance and computational complexity) among some machine learning based approaches used in QC for Bengali language.
2212.09448	Senem Tanberk PhD	Senem Tanberk, Mustafa Can	Smart Journey in Istanbul: A Mobile Application in Smart Cities for Traffic Estimation by Harnessing Time Series	null	null	10.1109/ASYU58738.2023.10296669	null	cs.AI cs.IR	http://creativecommons.org/licenses/by/4.0/	In recent decades, mobile applications (apps) have gained enormous popularity. Smart services for smart cities increasingly gain attention. The main goal of the proposed research is to present a new AI-powered mobile application on Istanbul's traffic congestion forecast by using traffic density data. It addresses the research question by using time series approaches (LSTM, Transformer, and XGBoost) based on past data over the traffic load dataset combined with meteorological conditions. Analysis of simulation results on predicted models will be discussed according to performance indicators such as MAPE, MAE, and RMSE. And then, it was observed that the Transformer model made the most accurate traffic prediction. The developed traffic forecasting prototype is expected to be a starting point on future products for a mobile application suitable for citizens' daily use.	[ { "created": "Tue, 13 Dec 2022 12:10:52 GMT", "version": "v1" } ]	2023-11-09	[ [ "Tanberk", "Senem", "" ], [ "Can", "Mustafa", "" ] ]	In recent decades, mobile applications (apps) have gained enormous popularity. Smart services for smart cities increasingly gain attention. The main goal of the proposed research is to present a new AI-powered mobile application on Istanbul's traffic congestion forecast by using traffic density data. It addresses the research question by using time series approaches (LSTM, Transformer, and XGBoost) based on past data over the traffic load dataset combined with meteorological conditions. Analysis of simulation results on predicted models will be discussed according to performance indicators such as MAPE, MAE, and RMSE. And then, it was observed that the Transformer model made the most accurate traffic prediction. The developed traffic forecasting prototype is expected to be a starting point on future products for a mobile application suitable for citizens' daily use.
2401.11150	Junxiao Shen Mr	Junxiao Shen, Xuhai Xu, Ran Tan, Amy Karlson, Evan Strasnick	Simultaneous Gesture Classification and Localization with an Automatic Gesture Annotation Model	null	null	null	null	cs.CV	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Training a real-time gesture recognition model heavily relies on annotated data. However, manual data annotation is costly and demands substantial human effort. In order to address this challenge, we propose a novel annotation model that can automatically annotate gesture classes and identify their temporal ranges. Our ablation study demonstrates that our annotation model design surpasses the baseline in terms of both gesture classification accuracy (3-4\% improvement) and localization accuracy (71-75\% improvement). We believe that this annotation model has immense potential to improve the training of downstream gesture recognition models using unlabeled datasets.	[ { "created": "Sat, 20 Jan 2024 07:11:03 GMT", "version": "v1" } ]	2024-01-23	[ [ "Shen", "Junxiao", "" ], [ "Xu", "Xuhai", "" ], [ "Tan", "Ran", "" ], [ "Karlson", "Amy", "" ], [ "Strasnick", "Evan", "" ] ]	Training a real-time gesture recognition model heavily relies on annotated data. However, manual data annotation is costly and demands substantial human effort. In order to address this challenge, we propose a novel annotation model that can automatically annotate gesture classes and identify their temporal ranges. Our ablation study demonstrates that our annotation model design surpasses the baseline in terms of both gesture classification accuracy (3-4\% improvement) and localization accuracy (71-75\% improvement). We believe that this annotation model has immense potential to improve the training of downstream gesture recognition models using unlabeled datasets.
1005.0907	Rdv Ijcsis	Yasser M. Alginaih, Abdul Ahad Siddiqi	Multistage Hybrid Arabic/Indian Numeral OCR System	IEEE Publication format, International Journal of Computer Science and Information Security, IJCSIS, Vol. 8 No. 1, April 2010, USA. ISSN 1947 5500, http://sites.google.com/site/ijcsis/	null	null	null	cs.CV	http://creativecommons.org/licenses/by-nc-sa/3.0/	The use of OCR in postal services is not yet universal and there are still many countries that process mail sorting manually. Automated Arabic/Indian numeral Optical Character Recognition (OCR) systems for Postal services are being used in some countries, but still there are errors during the mail sorting process, thus causing a reduction in efficiency. The need to investigate fast and efficient recognition algorithms/systems is important so as to correctly read the postal codes from mail addresses and to eliminate any errors during the mail sorting stage. The objective of this study is to recognize printed numerical postal codes from mail addresses. The proposed system is a multistage hybrid system which consists of three different feature extraction methods, i.e., binary, zoning, and fuzzy features, and three different classifiers, i.e., Hamming Nets, Euclidean Distance, and Fuzzy Neural Network Classifiers. The proposed system, systematically compares the performance of each of these methods, and ensures that the numerals are recognized correctly. Comprehensive results provide a very high recognition rate, outperforming the other known developed methods in literature.	[ { "created": "Thu, 6 May 2010 07:25:23 GMT", "version": "v1" } ]	2010-05-07	[ [ "Alginaih", "Yasser M.", "" ], [ "Siddiqi", "Abdul Ahad", "" ] ]	The use of OCR in postal services is not yet universal and there are still many countries that process mail sorting manually. Automated Arabic/Indian numeral Optical Character Recognition (OCR) systems for Postal services are being used in some countries, but still there are errors during the mail sorting process, thus causing a reduction in efficiency. The need to investigate fast and efficient recognition algorithms/systems is important so as to correctly read the postal codes from mail addresses and to eliminate any errors during the mail sorting stage. The objective of this study is to recognize printed numerical postal codes from mail addresses. The proposed system is a multistage hybrid system which consists of three different feature extraction methods, i.e., binary, zoning, and fuzzy features, and three different classifiers, i.e., Hamming Nets, Euclidean Distance, and Fuzzy Neural Network Classifiers. The proposed system, systematically compares the performance of each of these methods, and ensures that the numerals are recognized correctly. Comprehensive results provide a very high recognition rate, outperforming the other known developed methods in literature.
2405.13937	Xingtong Yu	Xingtong Yu, Zhenghao Liu, Yuan Fang, Xinming Zhang	DyGPrompt: Learning Feature and Time Prompts on Dynamic Graphs	Under review	null	null	null	cs.LG	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Dynamic graphs are pervasive in the real world, modeling dynamic relations between objects across various fields. For dynamic graph modeling, dynamic graph neural networks (DGNNs) have emerged as a mainstream technique, which are generally pre-trained on the link prediction task, leaving a significant gap from the objectives of downstream tasks such as node classification. To bridge the gap, prompt-based learning has gained traction on graphs. However, existing efforts focus on static graphs, neglecting the evolution of dynamic graphs. In this paper, we propose DyGPrompt, a novel pre-training and prompting framework for dynamic graph modeling. First, we design dual prompts to address the gap in both task objectives and dynamic variations across pre-training and downstream tasks. Second, we recognize that node and time features mutually characterize each other, and propose dual condition-nets to model the evolving node-time patterns in downstream tasks. Finally, we thoroughly evaluate and analyze DyGPrompt through extensive experiments on three public datasets.	[ { "created": "Wed, 22 May 2024 19:10:24 GMT", "version": "v1" }, { "created": "Sun, 26 May 2024 01:46:11 GMT", "version": "v2" }, { "created": "Tue, 28 May 2024 10:07:29 GMT", "version": "v3" }, { "created": "Tue, 2 Jul 2024 05:14:10 GMT", "version": "v4" }, { "created": "Wed, 3 Jul 2024 02:06:07 GMT", "version": "v5" } ]	2024-07-04	[ [ "Yu", "Xingtong", "" ], [ "Liu", "Zhenghao", "" ], [ "Fang", "Yuan", "" ], [ "Zhang", "Xinming", "" ] ]	Dynamic graphs are pervasive in the real world, modeling dynamic relations between objects across various fields. For dynamic graph modeling, dynamic graph neural networks (DGNNs) have emerged as a mainstream technique, which are generally pre-trained on the link prediction task, leaving a significant gap from the objectives of downstream tasks such as node classification. To bridge the gap, prompt-based learning has gained traction on graphs. However, existing efforts focus on static graphs, neglecting the evolution of dynamic graphs. In this paper, we propose DyGPrompt, a novel pre-training and prompting framework for dynamic graph modeling. First, we design dual prompts to address the gap in both task objectives and dynamic variations across pre-training and downstream tasks. Second, we recognize that node and time features mutually characterize each other, and propose dual condition-nets to model the evolving node-time patterns in downstream tasks. Finally, we thoroughly evaluate and analyze DyGPrompt through extensive experiments on three public datasets.
2208.02884	Nicolaas Kaashoek	Nicolaas Kaashoek and Robert Morris	CheckSync: Using Runtime-Integrated Checkpoints to Achieve High Availability}	14 pages, 6 figures	null	null	null	cs.DC	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	CheckSync provides applications with high availability via runtime-integrated checkpointing. This allows CheckSync to take checkpoints of a process running in a memory-managed language (Go, for now), which can be resumed on another machine after a failure. CheckSync uses the runtime to checkpoint only the process' live memory, doing without requiring significant changes to applications. CheckSync maintains the ease of use provided by virtual machines for the applications it supports without requiring that an entire virtual machine image be snapshotted. Because CheckSync captures only the memory used by an application, it produces checkpoints that are smaller (by an order of magnitude) than virtual machine snapshots if the memory footprint of the application is relatively small compared to the state of the rest of the operating system. Additionally, when running go-cache, a popular in-memory key/value store, CheckSync reduces throughput by only 12% compared to the 78% throughput loss when using go-cache's snapshot functionality, the 45% loss when using CRIU, and the 68% loss when using virtual machine live migration.	[ { "created": "Thu, 4 Aug 2022 20:53:50 GMT", "version": "v1" } ]	2022-08-08	[ [ "Kaashoek", "Nicolaas", "" ], [ "Morris", "Robert", "" ] ]	CheckSync provides applications with high availability via runtime-integrated checkpointing. This allows CheckSync to take checkpoints of a process running in a memory-managed language (Go, for now), which can be resumed on another machine after a failure. CheckSync uses the runtime to checkpoint only the process' live memory, doing without requiring significant changes to applications. CheckSync maintains the ease of use provided by virtual machines for the applications it supports without requiring that an entire virtual machine image be snapshotted. Because CheckSync captures only the memory used by an application, it produces checkpoints that are smaller (by an order of magnitude) than virtual machine snapshots if the memory footprint of the application is relatively small compared to the state of the rest of the operating system. Additionally, when running go-cache, a popular in-memory key/value store, CheckSync reduces throughput by only 12% compared to the 78% throughput loss when using go-cache's snapshot functionality, the 45% loss when using CRIU, and the 68% loss when using virtual machine live migration.
2303.13299	Avi Schwarzschild	Avi Schwarzschild, Max Cembalest, Karthik Rao, Keegan Hines, John Dickerson	Reckoning with the Disagreement Problem: Explanation Consensus as a Training Objective	null	null	null	null	cs.LG cs.AI	http://creativecommons.org/licenses/by/4.0/	As neural networks increasingly make critical decisions in high-stakes settings, monitoring and explaining their behavior in an understandable and trustworthy manner is a necessity. One commonly used type of explainer is post hoc feature attribution, a family of methods for giving each feature in an input a score corresponding to its influence on a model's output. A major limitation of this family of explainers in practice is that they can disagree on which features are more important than others. Our contribution in this paper is a method of training models with this disagreement problem in mind. We do this by introducing a Post hoc Explainer Agreement Regularization (PEAR) loss term alongside the standard term corresponding to accuracy, an additional term that measures the difference in feature attribution between a pair of explainers. We observe on three datasets that we can train a model with this loss term to improve explanation consensus on unseen data, and see improved consensus between explainers other than those used in the loss term. We examine the trade-off between improved consensus and model performance. And finally, we study the influence our method has on feature attribution explanations.	[ { "created": "Thu, 23 Mar 2023 14:35:37 GMT", "version": "v1" } ]	2023-03-24	[ [ "Schwarzschild", "Avi", "" ], [ "Cembalest", "Max", "" ], [ "Rao", "Karthik", "" ], [ "Hines", "Keegan", "" ], [ "Dickerson", "John", "" ] ]	As neural networks increasingly make critical decisions in high-stakes settings, monitoring and explaining their behavior in an understandable and trustworthy manner is a necessity. One commonly used type of explainer is post hoc feature attribution, a family of methods for giving each feature in an input a score corresponding to its influence on a model's output. A major limitation of this family of explainers in practice is that they can disagree on which features are more important than others. Our contribution in this paper is a method of training models with this disagreement problem in mind. We do this by introducing a Post hoc Explainer Agreement Regularization (PEAR) loss term alongside the standard term corresponding to accuracy, an additional term that measures the difference in feature attribution between a pair of explainers. We observe on three datasets that we can train a model with this loss term to improve explanation consensus on unseen data, and see improved consensus between explainers other than those used in the loss term. We examine the trade-off between improved consensus and model performance. And finally, we study the influence our method has on feature attribution explanations.
2311.11596	Yining Miao	Yining Miao, Nanlin Shi, Changxing Huang, Yonghao Song, Xiaogang Chen, Yijun Wang, Xiaorong Gao	High-performance cVEP-BCI under minimal calibration	35 pages, 5 figures	null	null	null	cs.HC cs.IT eess.SP math.IT q-bio.NC	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	The ultimate goal of brain-computer interfaces (BCIs) based on visual modulation paradigms is to achieve high-speed performance without the burden of extensive calibration. Code-modulated visual evoked potential-based BCIs (cVEP-BCIs) modulated by broadband white noise (WN) offer various advantages, including increased communication speed, expanded encoding target capabilities, and enhanced coding flexibility. However, the complexity of the spatial-temporal patterns under broadband stimuli necessitates extensive calibration for effective target identification in cVEP-BCIs. Consequently, the information transfer rate (ITR) of cVEP-BCI under limited calibration usually stays around 100 bits per minute (bpm), significantly lagging behind state-of-the-art steady-state visual evoked potential-based BCIs (SSVEP-BCIs), which achieve rates above 200 bpm. To enhance the performance of cVEP-BCIs with minimal calibration, we devised an efficient calibration stage involving a brief single-target flickering, lasting less than a minute, to extract generalizable spatial-temporal patterns. Leveraging the calibration data, we developed two complementary methods to construct cVEP temporal patterns: the linear modeling method based on the stimulus sequence and the transfer learning techniques using cross-subject data. As a result, we achieved the highest ITR of 250 bpm under a minute of calibration, which has been shown to be comparable to the state-of-the-art SSVEP paradigms. In summary, our work significantly improved the cVEP performance under few-shot learning, which is expected to expand the practicality and usability of cVEP-BCIs.	[ { "created": "Mon, 20 Nov 2023 08:20:51 GMT", "version": "v1" } ]	2023-11-21	[ [ "Miao", "Yining", "" ], [ "Shi", "Nanlin", "" ], [ "Huang", "Changxing", "" ], [ "Song", "Yonghao", "" ], [ "Chen", "Xiaogang", "" ], [ "Wang", "Yijun", "" ], [ "Gao", "Xiaorong", "" ] ]	The ultimate goal of brain-computer interfaces (BCIs) based on visual modulation paradigms is to achieve high-speed performance without the burden of extensive calibration. Code-modulated visual evoked potential-based BCIs (cVEP-BCIs) modulated by broadband white noise (WN) offer various advantages, including increased communication speed, expanded encoding target capabilities, and enhanced coding flexibility. However, the complexity of the spatial-temporal patterns under broadband stimuli necessitates extensive calibration for effective target identification in cVEP-BCIs. Consequently, the information transfer rate (ITR) of cVEP-BCI under limited calibration usually stays around 100 bits per minute (bpm), significantly lagging behind state-of-the-art steady-state visual evoked potential-based BCIs (SSVEP-BCIs), which achieve rates above 200 bpm. To enhance the performance of cVEP-BCIs with minimal calibration, we devised an efficient calibration stage involving a brief single-target flickering, lasting less than a minute, to extract generalizable spatial-temporal patterns. Leveraging the calibration data, we developed two complementary methods to construct cVEP temporal patterns: the linear modeling method based on the stimulus sequence and the transfer learning techniques using cross-subject data. As a result, we achieved the highest ITR of 250 bpm under a minute of calibration, which has been shown to be comparable to the state-of-the-art SSVEP paradigms. In summary, our work significantly improved the cVEP performance under few-shot learning, which is expected to expand the practicality and usability of cVEP-BCIs.
2204.08970	Zhihao Li	Zhihao Li, Si Yi, Zhan Ma	Rendering Nighttime Image Via Cascaded Color and Brightness Compensation	Accepted by NTIRE 2022 (CVPR Workshop)	null	null	null	cs.CV eess.IV	http://creativecommons.org/licenses/by/4.0/	Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website: https://njuvision.github.io/CBUnet.	[ { "created": "Tue, 19 Apr 2022 16:15:31 GMT", "version": "v1" }, { "created": "Thu, 21 Apr 2022 17:23:11 GMT", "version": "v2" } ]	2022-04-22	[ [ "Li", "Zhihao", "" ], [ "Yi", "Si", "" ], [ "Ma", "Zhan", "" ] ]	Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website: https://njuvision.github.io/CBUnet.
1702.05724	Daniel M\'endez Fern\'andez	Marco Kuhrmann and Daniel M\'endez Fern\'andez and Thomas Ternit\'e	On the Use of Variability Operations in the V-Modell XT Software Process Line	Journal of Software: Evolution and Process, 2015	null	10.1002/smr.1751	null	cs.SE	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Software process lines provide a systematic approach to develop and manage software processes. It defines a reference process containing general process assets, whereas a well-defined customization approach allows process engineers to create new process variants, e.g., by extending or modifying process assets. Variability operations are an instrument to realize flexibility by explicitly declaring required modifications, which are applied to create a procedurally generated company-specific process. However, little is known about which variability operations are suitable in practice. In this article, we present a study on the feasibility of variability operations to support the development of software process lines in the context of the V-Modell XT. We analyze which variability operations are defined and practically used. We provide an initial catalog of variability operations as an improvement proposal for other process models. Our findings show that 69 variability operation types are defined across several metamodel versions of which, however, 25 remain unused. The found variability operations allow for systematically modifying the content of process model elements and the process documentation, and they allow for altering the structure of a process model and its description. Furthermore, we also find that variability operations can help process engineers to compensate process metamodel evolution.	[ { "created": "Sun, 19 Feb 2017 08:58:12 GMT", "version": "v1" } ]	2017-02-21	[ [ "Kuhrmann", "Marco", "" ], [ "Fernández", "Daniel Méndez", "" ], [ "Ternité", "Thomas", "" ] ]	Software process lines provide a systematic approach to develop and manage software processes. It defines a reference process containing general process assets, whereas a well-defined customization approach allows process engineers to create new process variants, e.g., by extending or modifying process assets. Variability operations are an instrument to realize flexibility by explicitly declaring required modifications, which are applied to create a procedurally generated company-specific process. However, little is known about which variability operations are suitable in practice. In this article, we present a study on the feasibility of variability operations to support the development of software process lines in the context of the V-Modell XT. We analyze which variability operations are defined and practically used. We provide an initial catalog of variability operations as an improvement proposal for other process models. Our findings show that 69 variability operation types are defined across several metamodel versions of which, however, 25 remain unused. The found variability operations allow for systematically modifying the content of process model elements and the process documentation, and they allow for altering the structure of a process model and its description. Furthermore, we also find that variability operations can help process engineers to compensate process metamodel evolution.
1711.11017	Ethan Perez	Simon Brodeur, Ethan Perez, Ankesh Anand, Florian Golemo, Luca Celotti, Florian Strub, Jean Rouat, Hugo Larochelle, Aaron Courville	HoME: a Household Multimodal Environment	Presented at NIPS 2017's Visually-Grounded Interaction and Language Workshop	null	null	null	cs.AI cs.CL cs.CV cs.RO cs.SD eess.AS	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	We introduce HoME: a Household Multimodal Environment for artificial agents to learn from vision, audio, semantics, physics, and interaction with objects and other agents, all within a realistic context. HoME integrates over 45,000 diverse 3D house layouts based on the SUNCG dataset, a scale which may facilitate learning, generalization, and transfer. HoME is an open-source, OpenAI Gym-compatible platform extensible to tasks in reinforcement learning, language grounding, sound-based navigation, robotics, multi-agent learning, and more. We hope HoME better enables artificial agents to learn as humans do: in an interactive, multimodal, and richly contextualized setting.	[ { "created": "Wed, 29 Nov 2017 18:45:59 GMT", "version": "v1" } ]	2017-11-30	[ [ "Brodeur", "Simon", "" ], [ "Perez", "Ethan", "" ], [ "Anand", "Ankesh", "" ], [ "Golemo", "Florian", "" ], [ "Celotti", "Luca", "" ], [ "Strub", "Florian", "" ], [ "Rouat", "Jean", "" ], [ "Larochelle", "Hugo", "" ], [ "Courville", "Aaron", "" ] ]	We introduce HoME: a Household Multimodal Environment for artificial agents to learn from vision, audio, semantics, physics, and interaction with objects and other agents, all within a realistic context. HoME integrates over 45,000 diverse 3D house layouts based on the SUNCG dataset, a scale which may facilitate learning, generalization, and transfer. HoME is an open-source, OpenAI Gym-compatible platform extensible to tasks in reinforcement learning, language grounding, sound-based navigation, robotics, multi-agent learning, and more. We hope HoME better enables artificial agents to learn as humans do: in an interactive, multimodal, and richly contextualized setting.
1709.03421	Riccardo Sven Risuleo	Riccardo Sven Risuleo, Giulio Bottegal, H{\aa}kan Hjalmarsson	Modeling and identification of uncertain-input systems	27 Pages, submitted to Automatica	null	null	null	cs.SY	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	In this work, we present a new class of models, called uncertain-input models, that allows us to treat system-identification problems in which a linear system is subject to a partially unknown input signal. To encode prior information about the input or the linear system, we use Gaussian-process models. We estimate the model from data using the empirical Bayes approach: the input and the impulse responses of the linear system are estimated using the posterior means of the Gaussian-process models given the data, and the hyperparameters that characterize the Gaussian-process models are estimated from the marginal likelihood of the data. We propose an iterative algorithm to find the hyperparameters that relies on the EM method and results in simple update steps. In the most general formulation, neither the marginal likelihood nor the posterior distribution of the unknowns is tractable. Therefore, we propose two approximation approaches, one based on Markov-chain Monte Carlo techniques and one based on variational Bayes approximation. We also show special model structures for which the distributions are treatable exactly. Through numerical simulations, we study the application of the uncertain-input model to the identification of Hammerstein systems and cascaded linear systems. As part of the contribution of the paper, we show that this model structure encompasses many classical problems in system identification such as classical PEM, Hammerstein models, errors-in-variables problems, blind system identification, and cascaded linear systems. This allows us to build a systematic procedure to apply the algorithms proposed in this work to a wide class of classical problems.	[ { "created": "Mon, 11 Sep 2017 14:53:38 GMT", "version": "v1" } ]	2017-09-12	[ [ "Risuleo", "Riccardo Sven", "" ], [ "Bottegal", "Giulio", "" ], [ "Hjalmarsson", "Håkan", "" ] ]	In this work, we present a new class of models, called uncertain-input models, that allows us to treat system-identification problems in which a linear system is subject to a partially unknown input signal. To encode prior information about the input or the linear system, we use Gaussian-process models. We estimate the model from data using the empirical Bayes approach: the input and the impulse responses of the linear system are estimated using the posterior means of the Gaussian-process models given the data, and the hyperparameters that characterize the Gaussian-process models are estimated from the marginal likelihood of the data. We propose an iterative algorithm to find the hyperparameters that relies on the EM method and results in simple update steps. In the most general formulation, neither the marginal likelihood nor the posterior distribution of the unknowns is tractable. Therefore, we propose two approximation approaches, one based on Markov-chain Monte Carlo techniques and one based on variational Bayes approximation. We also show special model structures for which the distributions are treatable exactly. Through numerical simulations, we study the application of the uncertain-input model to the identification of Hammerstein systems and cascaded linear systems. As part of the contribution of the paper, we show that this model structure encompasses many classical problems in system identification such as classical PEM, Hammerstein models, errors-in-variables problems, blind system identification, and cascaded linear systems. This allows us to build a systematic procedure to apply the algorithms proposed in this work to a wide class of classical problems.
2110.04678	Ankit Parag Shah	Rita Singh, Ankit Shah, Hira Dhamyal	An Overview of Techniques for Biomarker Discovery in Voice Signal	Last two authors contributed equally to the paper	null	null	null	cs.SD cs.AI eess.AS	http://creativecommons.org/licenses/by-nc-nd/4.0/	This paper reflects on the effect of several categories of medical conditions on human voice, focusing on those that may be hypothesized to have effects on voice, but for which the changes themselves may be subtle enough to have eluded observation in standard analytical examinations of the voice signal. It presents three categories of techniques that can potentially uncover such elusive biomarkers and allow them to be measured and used for predictive and diagnostic purposes. These approaches include proxy techniques, model-based analytical techniques and data-driven AI techniques.	[ { "created": "Sun, 10 Oct 2021 01:39:28 GMT", "version": "v1" } ]	2021-10-12	[ [ "Singh", "Rita", "" ], [ "Shah", "Ankit", "" ], [ "Dhamyal", "Hira", "" ] ]	This paper reflects on the effect of several categories of medical conditions on human voice, focusing on those that may be hypothesized to have effects on voice, but for which the changes themselves may be subtle enough to have eluded observation in standard analytical examinations of the voice signal. It presents three categories of techniques that can potentially uncover such elusive biomarkers and allow them to be measured and used for predictive and diagnostic purposes. These approaches include proxy techniques, model-based analytical techniques and data-driven AI techniques.
2308.01286	Diptapriyo Majumdar	Diptapriyo Majumdar	Enumeration Kernels of Polynomial Size for Cuts of Bounded Degree	There have been major revision in the technicalities and proofs of the paper	null	null	null	cs.DS cs.DM	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	Enumeration kernelization was first proposed by Creignou et al. [TOCS 2017] and was later refined by Golovach et al. [JCSS 2022] into two different variants: fully-polynomial enumeration kernelization and polynomial-delay enumeration kernelization. In this paper, we consider the DEGREE-d-CUT problem from the perspective of (polynomial-delay) enumeration kenrelization. Given an undirected graph G = (V, E), a cut F = (A, B) is a degree-d-cut of G if every $u \in A$ has at most d neighbors in B and every $v \in B$ has at most d neighbors in A. Checking the existence of a degree-d-cut in a graph is a well-known NP-hard problem and is well-studied in parameterized complexity [Algorithmica 2021, IWOCA 2021]. This problem also generalizes a well-studied problem MATCHING CUT (set d = 1) that has been a central problem in the literature of polynomial-delay enumeration kernelization. In this paper, we study three different enumeration variants of this problem, ENUM DEGREE-d-CUT, ENUM MIN-DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT that intends to enumerate all the d-cuts, all the minimal d-cuts and all the maximal degree-d-cuts respectively. We consider various structural parameters of the input and for every fixed $d \geq 1$, we provide polynomial-delay enumeration kernelizations of polynomial size for ENUM DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT and fully-polynomial enumeration kernels of polynomial size for ENUM MIN-DEGREE-d-CUT.	[ { "created": "Wed, 2 Aug 2023 17:18:19 GMT", "version": "v1" }, { "created": "Thu, 30 Nov 2023 09:14:39 GMT", "version": "v2" }, { "created": "Fri, 1 Dec 2023 07:32:28 GMT", "version": "v3" }, { "created": "Fri, 2 Feb 2024 18:15:47 GMT", "version": "v4" }, { "created": "Sat, 27 Apr 2024 17:42:26 GMT", "version": "v5" } ]	2024-04-30	[ [ "Majumdar", "Diptapriyo", "" ] ]	Enumeration kernelization was first proposed by Creignou et al. [TOCS 2017] and was later refined by Golovach et al. [JCSS 2022] into two different variants: fully-polynomial enumeration kernelization and polynomial-delay enumeration kernelization. In this paper, we consider the DEGREE-d-CUT problem from the perspective of (polynomial-delay) enumeration kenrelization. Given an undirected graph G = (V, E), a cut F = (A, B) is a degree-d-cut of G if every $u \in A$ has at most d neighbors in B and every $v \in B$ has at most d neighbors in A. Checking the existence of a degree-d-cut in a graph is a well-known NP-hard problem and is well-studied in parameterized complexity [Algorithmica 2021, IWOCA 2021]. This problem also generalizes a well-studied problem MATCHING CUT (set d = 1) that has been a central problem in the literature of polynomial-delay enumeration kernelization. In this paper, we study three different enumeration variants of this problem, ENUM DEGREE-d-CUT, ENUM MIN-DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT that intends to enumerate all the d-cuts, all the minimal d-cuts and all the maximal degree-d-cuts respectively. We consider various structural parameters of the input and for every fixed $d \geq 1$, we provide polynomial-delay enumeration kernelizations of polynomial size for ENUM DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT and fully-polynomial enumeration kernels of polynomial size for ENUM MIN-DEGREE-d-CUT.
0810.2529	Alireza Bayesteh	Jamshid Abouei, Alireza Bayesteh, Masoud Ebrahimi, and Amir K. Khandani	On the Throughput Maximization in Dencentralized Wireless Networks	Submitted to IEEE Transactions on Information Theory	null	null	null	cs.IT math.IT	http://arxiv.org/licenses/nonexclusive-distrib/1.0/	A distributed single-hop wireless network with $K$ links is considered, where the links are partitioned into a fixed number ($M$) of clusters each operating in a subchannel with bandwidth $\frac{W}{M}$. The subchannels are assumed to be orthogonal to each other. A general shadow-fading model, described by parameters $(\alpha,\varpi)$, is considered where $\alpha$ denotes the probability of shadowing and $\varpi$ ($\varpi \leq 1$) represents the average cross-link gains. The main goal of this paper is to find the maximum network throughput in the asymptotic regime of $K \to \infty$, which is achieved by: i) proposing a distributed and non-iterative power allocation strategy, where the objective of each user is to maximize its best estimate (based on its local information, i.e., direct channel gain) of the average network throughput, and ii) choosing the optimum value for $M$. In the first part of the paper, the network hroughput is defined as the \textit{average sum-rate} of the network, which is shown to scale as $\Theta (\log K)$. Moreover, it is proved that in the strong interference scenario, the optimum power allocation strategy for each user is a threshold-based on-off scheme. In the second part, the network throughput is defined as the \textit{guaranteed sum-rate}, when the outage probability approaches zero. In this scenario, it is demonstrated that the on-off power allocation scheme maximizes the throughput, which scales as $\frac{W}{\alpha \varpi} \log K$. Moreover, the optimum spectrum sharing for maximizing the average sum-rate and the guaranteed sum-rate is achieved at M=1.	[ { "created": "Tue, 14 Oct 2008 19:40:22 GMT", "version": "v1" } ]	2008-10-15	[ [ "Abouei", "Jamshid", "" ], [ "Bayesteh", "Alireza", "" ], [ "Ebrahimi", "Masoud", "" ], [ "Khandani", "Amir K.", "" ] ]	A distributed single-hop wireless network with $K$ links is considered, where the links are partitioned into a fixed number ($M$) of clusters each operating in a subchannel with bandwidth $\frac{W}{M}$. The subchannels are assumed to be orthogonal to each other. A general shadow-fading model, described by parameters $(\alpha,\varpi)$, is considered where $\alpha$ denotes the probability of shadowing and $\varpi$ ($\varpi \leq 1$) represents the average cross-link gains. The main goal of this paper is to find the maximum network throughput in the asymptotic regime of $K \to \infty$, which is achieved by: i) proposing a distributed and non-iterative power allocation strategy, where the objective of each user is to maximize its best estimate (based on its local information, i.e., direct channel gain) of the average network throughput, and ii) choosing the optimum value for $M$. In the first part of the paper, the network hroughput is defined as the \textit{average sum-rate} of the network, which is shown to scale as $\Theta (\log K)$. Moreover, it is proved that in the strong interference scenario, the optimum power allocation strategy for each user is a threshold-based on-off scheme. In the second part, the network throughput is defined as the \textit{guaranteed sum-rate}, when the outage probability approaches zero. In this scenario, it is demonstrated that the on-off power allocation scheme maximizes the throughput, which scales as $\frac{W}{\alpha \varpi} \log K$. Moreover, the optimum spectrum sharing for maximizing the average sum-rate and the guaranteed sum-rate is achieved at M=1.
2010.08660	Manas Gaur	Manas Gaur, Keyur Faldu, Amit Sheth	Semantics of the Black-Box: Can knowledge graphs help make deep learning systems more interpretable and explainable?	6 pages + references, 4 figures, Accepted to IEEE internet computing 2020	null	null	null	cs.AI	http://creativecommons.org/licenses/by/4.0/	The recent series of innovations in deep learning (DL) have shown enormous potential to impact individuals and society, both positively and negatively. The DL models utilizing massive computing power and enormous datasets have significantly outperformed prior historical benchmarks on increasingly difficult, well-defined research tasks across technology domains such as computer vision, natural language processing, signal processing, and human-computer interactions. However, the Black-Box nature of DL models and their over-reliance on massive amounts of data condensed into labels and dense representations poses challenges for interpretability and explainability of the system. Furthermore, DLs have not yet been proven in their ability to effectively utilize relevant domain knowledge and experience critical to human understanding. This aspect is missing in early data-focused approaches and necessitated knowledge-infused learning and other strategies to incorporate computational knowledge. This article demonstrates how knowledge, provided as a knowledge graph, is incorporated into DL methods using knowledge-infused learning, which is one of the strategies. We then discuss how this makes a fundamental difference in the interpretability and explainability of current approaches, and illustrate it with examples from natural language processing for healthcare and education applications.	[ { "created": "Fri, 16 Oct 2020 22:55:23 GMT", "version": "v1" }, { "created": "Sun, 1 Nov 2020 02:28:43 GMT", "version": "v2" }, { "created": "Tue, 3 Nov 2020 15:52:55 GMT", "version": "v3" }, { "created": "Fri, 11 Dec 2020 23:03:11 GMT", "version": "v4" } ]	2020-12-15	[ [ "Gaur", "Manas", "" ], [ "Faldu", "Keyur", "" ], [ "Sheth", "Amit", "" ] ]	The recent series of innovations in deep learning (DL) have shown enormous potential to impact individuals and society, both positively and negatively. The DL models utilizing massive computing power and enormous datasets have significantly outperformed prior historical benchmarks on increasingly difficult, well-defined research tasks across technology domains such as computer vision, natural language processing, signal processing, and human-computer interactions. However, the Black-Box nature of DL models and their over-reliance on massive amounts of data condensed into labels and dense representations poses challenges for interpretability and explainability of the system. Furthermore, DLs have not yet been proven in their ability to effectively utilize relevant domain knowledge and experience critical to human understanding. This aspect is missing in early data-focused approaches and necessitated knowledge-infused learning and other strategies to incorporate computational knowledge. This article demonstrates how knowledge, provided as a knowledge graph, is incorporated into DL methods using knowledge-infused learning, which is one of the strategies. We then discuss how this makes a fundamental difference in the interpretability and explainability of current approaches, and illustrate it with examples from natural language processing for healthcare and education applications.

1906.10370

Miguel Sepulcre

Rafael Molina-Masegosa, Miguel Sepulcre and Javier Gozalvez

Geo-Based Scheduling for C-V2X Networks

null

IEEE Transactions on Vehicular Technology, Volume 68, Issue 9, pp. 8397 - 8407, September 2019

10.1109/TVT.2019.2924698

null

cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Cellular Vehicle-to-Everything (C-V2X) networks can operate without cellular infrastructure support. Vehicles can autonomously select their radio resources using the sensing-based Semi-Persistent Scheduling (SPS) algorithm specified by the Third Generation Partnership Project (3GPP). The sensing nature of the SPS scheme makes C-V2X communications prone to the well-known hidden-terminal problem. To address this problem, this paper proposes a novel geo-based scheduling scheme that allows vehicles to autonomously select their radio resources based on the location and ordering of neighboring vehicles on the road. The proposed scheme results in an implicit resource selection coordination between vehicles (even with those outside the sensing range) that reduces packet collisions. This paper evaluates analytically and through simulations the proposed scheduling scheme. The obtained results demonstrate that it reduces packet collisions and significantly increases the C-V2X performance compared to when using the sensing-based SPS scheme.

[ { "created": "Tue, 25 Jun 2019 08:12:28 GMT", "version": "v1" } ]

2019-09-20

[ [ "Molina-Masegosa", "Rafael", "" ], [ "Sepulcre", "Miguel", "" ], [ "Gozalvez", "Javier", "" ] ]

Cellular Vehicle-to-Everything (C-V2X) networks can operate without cellular infrastructure support. Vehicles can autonomously select their radio resources using the sensing-based Semi-Persistent Scheduling (SPS) algorithm specified by the Third Generation Partnership Project (3GPP). The sensing nature of the SPS scheme makes C-V2X communications prone to the well-known hidden-terminal problem. To address this problem, this paper proposes a novel geo-based scheduling scheme that allows vehicles to autonomously select their radio resources based on the location and ordering of neighboring vehicles on the road. The proposed scheme results in an implicit resource selection coordination between vehicles (even with those outside the sensing range) that reduces packet collisions. This paper evaluates analytically and through simulations the proposed scheduling scheme. The obtained results demonstrate that it reduces packet collisions and significantly increases the C-V2X performance compared to when using the sensing-based SPS scheme.

2312.07340

Yusen Feng

Yusen Feng, Xiyan Xu, Libin Liu

MuscleVAE: Model-Based Controllers of Muscle-Actuated Characters

null

cs.GR

http://creativecommons.org/licenses/by/4.0/

In this paper, we present a simulation and control framework for generating biomechanically plausible motion for muscle-actuated characters. We incorporate a fatigue dynamics model, the 3CC-r model, into the widely-adopted Hill-type muscle model to simulate the development and recovery of fatigue in muscles, which creates a natural evolution of motion style caused by the accumulation of fatigue from prolonged activities. To address the challenging problem of controlling a musculoskeletal system with high degrees of freedom, we propose a novel muscle-space control strategy based on PD control. Our simulation and control framework facilitates the training of a generative model for muscle-based motion control, which we refer to as MuscleVAE. By leveraging the variational autoencoders (VAEs), MuscleVAE is capable of learning a rich and flexible latent representation of skills from a large unstructured motion dataset, encoding not only motion features but also muscle control and fatigue properties. We demonstrate that the MuscleVAE model can be efficiently trained using a model-based approach, resulting in the production of high-fidelity motions and enabling a variety of downstream tasks.

[ { "created": "Tue, 12 Dec 2023 15:01:17 GMT", "version": "v1" } ]

2023-12-13

[ [ "Feng", "Yusen", "" ], [ "Xu", "Xiyan", "" ], [ "Liu", "Libin", "" ] ]

In this paper, we present a simulation and control framework for generating biomechanically plausible motion for muscle-actuated characters. We incorporate a fatigue dynamics model, the 3CC-r model, into the widely-adopted Hill-type muscle model to simulate the development and recovery of fatigue in muscles, which creates a natural evolution of motion style caused by the accumulation of fatigue from prolonged activities. To address the challenging problem of controlling a musculoskeletal system with high degrees of freedom, we propose a novel muscle-space control strategy based on PD control. Our simulation and control framework facilitates the training of a generative model for muscle-based motion control, which we refer to as MuscleVAE. By leveraging the variational autoencoders (VAEs), MuscleVAE is capable of learning a rich and flexible latent representation of skills from a large unstructured motion dataset, encoding not only motion features but also muscle control and fatigue properties. We demonstrate that the MuscleVAE model can be efficiently trained using a model-based approach, resulting in the production of high-fidelity motions and enabling a variety of downstream tasks.

2211.08292

Jennifer Andreoli-Fang

Jennifer Andreoli-Fang, John T Chapman

Mobile-Aware Scheduling for Low Latency Backhaul over DOCSIS

IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), 2017

null

10.1109/PIMRC.2017.8292173

null

cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this paper, we discuss latency reduction techniques for mobile backhaul over Data Over Cable Service Interface Specifications (DOCSIS) networks. When the latencies from both the wireless and the DOCSIS networks are added together, it can result in noticeable end-to-end system latency, particularly under network congestion. Previously, we proposed a method to improve upstream user-to-mobile core latency by coordinating the LTE and DOCSIS scheduling. The method reduces the impact on system latency from the DOCSIS network's request-grant-data loop, which is the main contributor of backhaul upstream latency. Since the method reduces latency on the DOCSIS data path, it will therefore improve performance of latency sensitive applications, particularly if TCP is used as the transport protocol, especially when the link is congested. In this paper, we investigate the effect of HARQ failure on system performance. Through simulation, we show that despite the uncertainty introduced by the LTE protocol, coordinated scheduling improves overall system latency.

[ { "created": "Tue, 15 Nov 2022 16:45:26 GMT", "version": "v1" }, { "created": "Wed, 16 Nov 2022 16:38:44 GMT", "version": "v2" } ]

2022-11-17

[ [ "Andreoli-Fang", "Jennifer", "" ], [ "Chapman", "John T", "" ] ]

In this paper, we discuss latency reduction techniques for mobile backhaul over Data Over Cable Service Interface Specifications (DOCSIS) networks. When the latencies from both the wireless and the DOCSIS networks are added together, it can result in noticeable end-to-end system latency, particularly under network congestion. Previously, we proposed a method to improve upstream user-to-mobile core latency by coordinating the LTE and DOCSIS scheduling. The method reduces the impact on system latency from the DOCSIS network's request-grant-data loop, which is the main contributor of backhaul upstream latency. Since the method reduces latency on the DOCSIS data path, it will therefore improve performance of latency sensitive applications, particularly if TCP is used as the transport protocol, especially when the link is congested. In this paper, we investigate the effect of HARQ failure on system performance. Through simulation, we show that despite the uncertainty introduced by the LTE protocol, coordinated scheduling improves overall system latency.

2103.14856

Ciriaco Andrea D'Angelo

Giovanni Abramo, Ciriaco Andrea D'Angelo, Lin Zhang

A comparison of two approaches for measuring interdisciplinary research output: the disciplinary diversity of authors vs the disciplinary diversity of the reference list

null

Journal of Informetrics, 12(4), 2018, 1182-1193

10.1016/j.joi.2018.09.001

null

cs.DL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This study investigates the convergence of two bibliometric approaches to the measurement of interdisciplinary research: one based on analyzing disciplinary diversity in the reference list of publications, the other based on the disciplinary diversity of authors of publications. In particular we measure the variety, balance, disparity and integrated diversity index of, respectively, single-author, multi-author single-field, and multi-author multi-field publications. We find that, in general, the diversity of the reference list grows with the number of fields reflected in a paper's authors' list and, to a lesser extent, with the number of authors being equal the number of fields. Further, we find that when fields belonging to different disciplines are reflected in the authors' list, the disparity in the reference list is higher than in the case of fields belonging to the same discipline. However, this general tendency varies across disciplines, and noticeable exceptions are found at individual paper level.

[ { "created": "Sat, 27 Mar 2021 09:34:53 GMT", "version": "v1" } ]

2021-03-30

[ [ "Abramo", "Giovanni", "" ], [ "D'Angelo", "Ciriaco Andrea", "" ], [ "Zhang", "Lin", "" ] ]

This study investigates the convergence of two bibliometric approaches to the measurement of interdisciplinary research: one based on analyzing disciplinary diversity in the reference list of publications, the other based on the disciplinary diversity of authors of publications. In particular we measure the variety, balance, disparity and integrated diversity index of, respectively, single-author, multi-author single-field, and multi-author multi-field publications. We find that, in general, the diversity of the reference list grows with the number of fields reflected in a paper's authors' list and, to a lesser extent, with the number of authors being equal the number of fields. Further, we find that when fields belonging to different disciplines are reflected in the authors' list, the disparity in the reference list is higher than in the case of fields belonging to the same discipline. However, this general tendency varies across disciplines, and noticeable exceptions are found at individual paper level.

1910.12435

Marcel Keller

Anders Dalskov and Daniel Escudero and Marcel Keller

Secure Evaluation of Quantized Neural Networks

22 pages

Proceedings on Privacy Enhancing Technologies 4 (2020): 355-375

10.2478/popets-2020-0077

null

cs.CR cs.LG

http://creativecommons.org/licenses/by-nc-nd/4.0/

We investigate two questions in this paper: First, we ask to what extent "MPC friendly" models are already supported by major Machine Learning frameworks such as TensorFlow or PyTorch. Prior works provide protocols that only work on fixed-point integers and specialized activation functions, two aspects that are not supported by popular Machine Learning frameworks, and the need for these specialized model representations means that it is hard, and often impossible, to use e.g., TensorFlow to design, train and test models that later have to be evaluated securely. Second, we ask to what extent the functionality for evaluating Neural Networks already exists in general-purpose MPC frameworks. These frameworks have received more scrutiny, are better documented and supported on more platforms. Furthermore, they are typically flexible in terms of the threat model they support. In contrast, most secure evaluation protocols in the literature are targeted to a specific threat model and their implementations are only a "proof-of-concept", making it very hard for their adoption in practice. We answer both of the above questions in a positive way: We observe that the quantization techniques supported by both TensorFlow, PyTorch and MXNet can provide models in a representation that can be evaluated securely; and moreover, that this evaluation can be performed by a general purpose MPC framework. We perform extensive benchmarks to understand the exact trade-offs between different corruption models, network sizes and efficiency. These experiments provide an interesting insight into cost between active and passive security, as well as honest and dishonest majority. Our work shows then that the separating line between existing ML frameworks and existing MPC protocols may be narrower than implicitly suggested by previous works.

[ { "created": "Mon, 28 Oct 2019 04:17:33 GMT", "version": "v1" }, { "created": "Mon, 1 Mar 2021 04:22:07 GMT", "version": "v2" } ]

2021-03-02

[ [ "Dalskov", "Anders", "" ], [ "Escudero", "Daniel", "" ], [ "Keller", "Marcel", "" ] ]

We investigate two questions in this paper: First, we ask to what extent "MPC friendly" models are already supported by major Machine Learning frameworks such as TensorFlow or PyTorch. Prior works provide protocols that only work on fixed-point integers and specialized activation functions, two aspects that are not supported by popular Machine Learning frameworks, and the need for these specialized model representations means that it is hard, and often impossible, to use e.g., TensorFlow to design, train and test models that later have to be evaluated securely. Second, we ask to what extent the functionality for evaluating Neural Networks already exists in general-purpose MPC frameworks. These frameworks have received more scrutiny, are better documented and supported on more platforms. Furthermore, they are typically flexible in terms of the threat model they support. In contrast, most secure evaluation protocols in the literature are targeted to a specific threat model and their implementations are only a "proof-of-concept", making it very hard for their adoption in practice. We answer both of the above questions in a positive way: We observe that the quantization techniques supported by both TensorFlow, PyTorch and MXNet can provide models in a representation that can be evaluated securely; and moreover, that this evaluation can be performed by a general purpose MPC framework. We perform extensive benchmarks to understand the exact trade-offs between different corruption models, network sizes and efficiency. These experiments provide an interesting insight into cost between active and passive security, as well as honest and dishonest majority. Our work shows then that the separating line between existing ML frameworks and existing MPC protocols may be narrower than implicitly suggested by previous works.

2407.06855

Arnab Sharma

Sourabh Kapoor, Arnab Sharma, Michael R\"oder, Caglar Demir, Axel-Cyrille Ngonga Ngomo

Performance Evaluation of Knowledge Graph Embedding Approaches under Non-adversarial Attacks

null

cs.LG cs.CR

http://creativecommons.org/licenses/by/4.0/

Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.

[ { "created": "Tue, 9 Jul 2024 13:42:14 GMT", "version": "v1" } ]

2024-07-10

[ [ "Kapoor", "Sourabh", "" ], [ "Sharma", "Arnab", "" ], [ "Röder", "Michael", "" ], [ "Demir", "Caglar", "" ], [ "Ngomo", "Axel-Cyrille Ngonga", "" ] ]

Knowledge Graph Embedding (KGE) transforms a discrete Knowledge Graph (KG) into a continuous vector space facilitating its use in various AI-driven applications like Semantic Search, Question Answering, or Recommenders. While KGE approaches are effective in these applications, most existing approaches assume that all information in the given KG is correct. This enables attackers to influence the output of these approaches, e.g., by perturbing the input. Consequently, the robustness of such KGE approaches has to be addressed. Recent work focused on adversarial attacks. However, non-adversarial attacks on all attack surfaces of these approaches have not been thoroughly examined. We close this gap by evaluating the impact of non-adversarial attacks on the performance of 5 state-of-the-art KGE algorithms on 5 datasets with respect to attacks on 3 attack surfaces-graph, parameter, and label perturbation. Our evaluation results suggest that label perturbation has a strong effect on the KGE performance, followed by parameter perturbation with a moderate and graph with a low effect.

1211.1861

Mohamed Firdhous

Automating Legal Research through Data Mining

8 pages, 11 figures, published in (IJACSA) International Journal of Advanced Computer Science and Applications. arXiv admin note: text overlap with wikipedia entry on text mining

International Journal of Advanced Computer Science and Applications (IJACSA) Vol. 1, No 6, December 2010, pp. 9-16

null

cs.IR

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The term legal research generally refers to the process of identifying and retrieving appropriate information necessary to support legal decision making from past case records. At present, the process is mostly manual, but some traditional technologies such as keyword searching are commonly used to speed the process up. But a keyword search is not a comprehensive search to cater to the requirements of legal research as the search result includes too many false hits in terms of irrelevant case records. Hence the present generic tools cannot be used to automate legal research. This paper presents a framework which was developed by combining several Text Mining techniques to automate the process overcoming the difficulties in the existing methods. Further, the research also identifies the possible enhancements that could be done to enhance the effectiveness of the framework.

[ { "created": "Thu, 8 Nov 2012 14:19:49 GMT", "version": "v1" } ]

2012-11-09

[ [ "Firdhous", "Mohamed", "" ] ]

The term legal research generally refers to the process of identifying and retrieving appropriate information necessary to support legal decision making from past case records. At present, the process is mostly manual, but some traditional technologies such as keyword searching are commonly used to speed the process up. But a keyword search is not a comprehensive search to cater to the requirements of legal research as the search result includes too many false hits in terms of irrelevant case records. Hence the present generic tools cannot be used to automate legal research. This paper presents a framework which was developed by combining several Text Mining techniques to automate the process overcoming the difficulties in the existing methods. Further, the research also identifies the possible enhancements that could be done to enhance the effectiveness of the framework.

1602.00810

Jean-Guillaume Dumas

Jean-Guillaume Dumas (LJK), Erich Kaltofen (NCSU), Emmanuel Thom\'e (CARAMBA), Gilles Villard (ARIC, LIP)

Linear Time Interactive Certificates for the Minimal Polynomial and the Determinant of a Sparse Matrix

null

International Symposium on Symbolic and Algebraic Computation, Jul 2016, Waterloo, Canada. pp.199-206, \&\#x27E8;10.1145/2930889.2930908\&\#x27E9

null

cs.SC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Computational problem certificates are additional data structures for each output, which can be used by a-possibly randomized-verification algorithm that proves the correctness of each output. In this paper, we give an algorithm that computes a certificate for the minimal polynomial of sparse or structured nxn matrices over an abstract field, of sufficiently large cardinality, whose Monte Carlo verification complexity requires a single matrix-vector multiplication and a linear number of extra field operations. We also propose a novel preconditioner that ensures irreducibility of the characteristic polynomial of the generically preconditioned matrix. This preconditioner takes linear time to be applied and uses only two random entries. We then combine these two techniques to give algorithms that compute certificates for the determinant, and thus for the characteristic polynomial, whose Monte Carlo verification complexity is therefore also linear.

[ { "created": "Tue, 2 Feb 2016 07:29:28 GMT", "version": "v1" }, { "created": "Mon, 2 Dec 2019 13:02:25 GMT", "version": "v2" } ]

2019-12-03

[ [ "Dumas", "Jean-Guillaume", "", "LJK" ], [ "Kaltofen", "Erich", "", "NCSU" ], [ "Thomé", "Emmanuel", "", "CARAMBA" ], [ "Villard", "Gilles", "", "ARIC, LIP" ] ]

Computational problem certificates are additional data structures for each output, which can be used by a-possibly randomized-verification algorithm that proves the correctness of each output. In this paper, we give an algorithm that computes a certificate for the minimal polynomial of sparse or structured nxn matrices over an abstract field, of sufficiently large cardinality, whose Monte Carlo verification complexity requires a single matrix-vector multiplication and a linear number of extra field operations. We also propose a novel preconditioner that ensures irreducibility of the characteristic polynomial of the generically preconditioned matrix. This preconditioner takes linear time to be applied and uses only two random entries. We then combine these two techniques to give algorithms that compute certificates for the determinant, and thus for the characteristic polynomial, whose Monte Carlo verification complexity is therefore also linear.

1912.07976

Heng Yang

Heng Yang, Biqing Zeng, JianHao Yang, Youwei Song and Ruyang Xu

A Multi-task Learning Model for Chinese-oriented Aspect Polarity Classification and Aspect Term Extraction

Submitted to Elsevier

null

cs.CL cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Aspect-based sentiment analysis (ABSA) task is a multi-grained task of natural language processing and consists of two subtasks: aspect term extraction (ATE) and aspect polarity classification (APC). Most of the existing work focuses on the subtask of aspect term polarity inferring and ignores the significance of aspect term extraction. Besides, the existing researches do not pay attention to the research of the Chinese-oriented ABSA task. Based on the local context focus (LCF) mechanism, this paper firstly proposes a multi-task learning model for Chinese-oriented aspect-based sentiment analysis, namely LCF-ATEPC. Compared with existing models, this model equips the capability of extracting aspect term and inferring aspect term polarity synchronously, moreover, this model is effective to analyze both Chinese and English comments simultaneously and the experiment on a multilingual mixed dataset proved its availability. By integrating the domain-adapted BERT model, the LCF-ATEPC model achieved the state-of-the-art performance of aspect term extraction and aspect polarity classification in four Chinese review datasets. Besides, the experimental results on the most commonly used SemEval-2014 task4 Restaurant and Laptop datasets outperform the state-of-the-art performance on the ATE and APC subtask.

[ { "created": "Tue, 17 Dec 2019 12:47:33 GMT", "version": "v1" }, { "created": "Thu, 19 Dec 2019 01:38:38 GMT", "version": "v2" }, { "created": "Wed, 12 Feb 2020 09:20:28 GMT", "version": "v3" } ]

2020-02-13

[ [ "Yang", "Heng", "" ], [ "Zeng", "Biqing", "" ], [ "Yang", "JianHao", "" ], [ "Song", "Youwei", "" ], [ "Xu", "Ruyang", "" ] ]

Aspect-based sentiment analysis (ABSA) task is a multi-grained task of natural language processing and consists of two subtasks: aspect term extraction (ATE) and aspect polarity classification (APC). Most of the existing work focuses on the subtask of aspect term polarity inferring and ignores the significance of aspect term extraction. Besides, the existing researches do not pay attention to the research of the Chinese-oriented ABSA task. Based on the local context focus (LCF) mechanism, this paper firstly proposes a multi-task learning model for Chinese-oriented aspect-based sentiment analysis, namely LCF-ATEPC. Compared with existing models, this model equips the capability of extracting aspect term and inferring aspect term polarity synchronously, moreover, this model is effective to analyze both Chinese and English comments simultaneously and the experiment on a multilingual mixed dataset proved its availability. By integrating the domain-adapted BERT model, the LCF-ATEPC model achieved the state-of-the-art performance of aspect term extraction and aspect polarity classification in four Chinese review datasets. Besides, the experimental results on the most commonly used SemEval-2014 task4 Restaurant and Laptop datasets outperform the state-of-the-art performance on the ATE and APC subtask.

2007.00584

Guillaume Jaume

Pushpak Pati, Guillaume Jaume, Lauren Alisha Fernandes, Antonio Foncubierta, Florinda Feroce, Anna Maria Anniciello, Giosue Scognamiglio, Nadia Brancati, Daniel Riccio, Maurizio Do Bonito, Giuseppe De Pietro, Gerardo Botti, Orcun Goksel, Jean-Philippe Thiran, Maria Frucci, Maria Gabrani

HACT-Net: A Hierarchical Cell-to-Tissue Graph Neural Network for Histopathological Image Classification

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Cancer diagnosis, prognosis, and therapeutic response prediction are heavily influenced by the relationship between the histopathological structures and the function of the tissue. Recent approaches acknowledging the structure-function relationship, have linked the structural and spatial patterns of cell organization in tissue via cell-graphs to tumor grades. Though cell organization is imperative, it is insufficient to entirely represent the histopathological structure. We propose a novel hierarchical cell-to-tissue-graph (HACT) representation to improve the structural depiction of the tissue. It consists of a low-level cell-graph, capturing cell morphology and interactions, a high-level tissue-graph, capturing morphology and spatial distribution of tissue parts, and cells-to-tissue hierarchies, encoding the relative spatial distribution of the cells with respect to the tissue distribution. Further, a hierarchical graph neural network (HACT-Net) is proposed to efficiently map the HACT representations to histopathological breast cancer subtypes. We assess the methodology on a large set of annotated tissue regions of interest from H\&E stained breast carcinoma whole-slides. Upon evaluation, the proposed method outperformed recent convolutional neural network and graph neural network approaches for breast cancer multi-class subtyping. The proposed entity-based topological analysis is more inline with the pathological diagnostic procedure of the tissue. It provides more command over the tissue modelling, therefore encourages the further inclusion of pathological priors into task-specific tissue representation.

[ { "created": "Wed, 1 Jul 2020 16:22:48 GMT", "version": "v1" } ]

2020-07-02

[ [ "Pati", "Pushpak", "" ], [ "Jaume", "Guillaume", "" ], [ "Fernandes", "Lauren Alisha", "" ], [ "Foncubierta", "Antonio", "" ], [ "Feroce", "Florinda", "" ], [ "Anniciello", "Anna Maria", "" ], [ "Scognamiglio", "Giosue", "" ], [ "Brancati", "Nadia", "" ], [ "Riccio", "Daniel", "" ], [ "Bonito", "Maurizio Do", "" ], [ "De Pietro", "Giuseppe", "" ], [ "Botti", "Gerardo", "" ], [ "Goksel", "Orcun", "" ], [ "Thiran", "Jean-Philippe", "" ], [ "Frucci", "Maria", "" ], [ "Gabrani", "Maria", "" ] ]

Cancer diagnosis, prognosis, and therapeutic response prediction are heavily influenced by the relationship between the histopathological structures and the function of the tissue. Recent approaches acknowledging the structure-function relationship, have linked the structural and spatial patterns of cell organization in tissue via cell-graphs to tumor grades. Though cell organization is imperative, it is insufficient to entirely represent the histopathological structure. We propose a novel hierarchical cell-to-tissue-graph (HACT) representation to improve the structural depiction of the tissue. It consists of a low-level cell-graph, capturing cell morphology and interactions, a high-level tissue-graph, capturing morphology and spatial distribution of tissue parts, and cells-to-tissue hierarchies, encoding the relative spatial distribution of the cells with respect to the tissue distribution. Further, a hierarchical graph neural network (HACT-Net) is proposed to efficiently map the HACT representations to histopathological breast cancer subtypes. We assess the methodology on a large set of annotated tissue regions of interest from H\&E stained breast carcinoma whole-slides. Upon evaluation, the proposed method outperformed recent convolutional neural network and graph neural network approaches for breast cancer multi-class subtyping. The proposed entity-based topological analysis is more inline with the pathological diagnostic procedure of the tissue. It provides more command over the tissue modelling, therefore encourages the further inclusion of pathological priors into task-specific tissue representation.

0810.5057

Patricia Gautier

Claire Fran\c{c}ois (INIST), Jean-Charles Lamirel (INRIA Lorraine - LORIA), Shadi Al Shehabi (INRIA Lorraine - LORIA)

Combining Advanced Visualization and Automatized Reasoning for Webometrics: A Test Study

null

COLLNET 2006, France (2006)

null

cs.IR cs.DL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper presents a first attempt at performing a precise and automatic identification of the linking behaviour in a scientific domain through the analysis of the communication of the related academic institutions on the web. The proposed approach is based on the paradigm of multiple viewpoint data analysis (MVDA) than can be fruitfully exploited to highlight relationships between data, like websites, carrying several kinds of description. It uses the MultiSOM clustering and mapping method. The domain that has been chosen for this study is the domain of Computer Science in Germany. The analysis is conduced on a set of 438 websites of this domain using all together, thematic, geographic and linking information. It highlights interesting results concerning both global and local linking behaviour.

[ { "created": "Tue, 28 Oct 2008 15:43:45 GMT", "version": "v1" }, { "created": "Tue, 20 Oct 2009 15:27:41 GMT", "version": "v2" } ]

2009-10-20

[ [ "François", "Claire", "", "INIST" ], [ "Lamirel", "Jean-Charles", "", "INRIA Lorraine -\n LORIA" ], [ "Shehabi", "Shadi Al", "", "INRIA Lorraine - LORIA" ] ]

This paper presents a first attempt at performing a precise and automatic identification of the linking behaviour in a scientific domain through the analysis of the communication of the related academic institutions on the web. The proposed approach is based on the paradigm of multiple viewpoint data analysis (MVDA) than can be fruitfully exploited to highlight relationships between data, like websites, carrying several kinds of description. It uses the MultiSOM clustering and mapping method. The domain that has been chosen for this study is the domain of Computer Science in Germany. The analysis is conduced on a set of 438 websites of this domain using all together, thematic, geographic and linking information. It highlights interesting results concerning both global and local linking behaviour.

0808.3651

Lijun Zhang

Lijun Zhang, Holger Hermanns, Friedrich Eisenbrand and David N. Jansen

Flow Faster: Efficient Decision Algorithms for Probabilistic Simulations

LMCS

Logical Methods in Computer Science, Volume 4, Issue 4 (November 11, 2008) lmcs:989

10.2168/LMCS-4(4:6)2008

null

cs.LO

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Strong and weak simulation relations have been proposed for Markov chains, while strong simulation and strong probabilistic simulation relations have been proposed for probabilistic automata. However, decision algorithms for strong and weak simulation over Markov chains, and for strong simulation over probabilistic automata are not efficient, which makes it as yet unclear whether they can be used as effectively as their non-probabilistic counterparts. This paper presents drastically improved algorithms to decide whether some (discrete- or continuous-time) Markov chain strongly or weakly simulates another, or whether a probabilistic automaton strongly simulates another. The key innovation is the use of parametric maximum flow techniques to amortize computations. We also present a novel algorithm for deciding strong probabilistic simulation preorders on probabilistic automata, which has polynomial complexity via a reduction to an LP problem. When extending the algorithms for probabilistic automata to their continuous-time counterpart, we retain the same complexity for both strong and strong probabilistic simulations.

[ { "created": "Wed, 27 Aug 2008 08:35:44 GMT", "version": "v1" }, { "created": "Mon, 10 Nov 2008 23:56:01 GMT", "version": "v2" }, { "created": "Tue, 18 Nov 2008 17:00:30 GMT", "version": "v3" } ]

2015-07-01

[ [ "Zhang", "Lijun", "" ], [ "Hermanns", "Holger", "" ], [ "Eisenbrand", "Friedrich", "" ], [ "Jansen", "David N.", "" ] ]

Strong and weak simulation relations have been proposed for Markov chains, while strong simulation and strong probabilistic simulation relations have been proposed for probabilistic automata. However, decision algorithms for strong and weak simulation over Markov chains, and for strong simulation over probabilistic automata are not efficient, which makes it as yet unclear whether they can be used as effectively as their non-probabilistic counterparts. This paper presents drastically improved algorithms to decide whether some (discrete- or continuous-time) Markov chain strongly or weakly simulates another, or whether a probabilistic automaton strongly simulates another. The key innovation is the use of parametric maximum flow techniques to amortize computations. We also present a novel algorithm for deciding strong probabilistic simulation preorders on probabilistic automata, which has polynomial complexity via a reduction to an LP problem. When extending the algorithms for probabilistic automata to their continuous-time counterpart, we retain the same complexity for both strong and strong probabilistic simulations.

2403.06757

Anthony Frion

Anthony Frion, Lucas Drumetz, Guillaume Tochon, Mauro Dalla Mura, Albdeldjalil A\"issa El Bey

Koopman Ensembles for Probabilistic Time Series Forecasting

null

cs.LG

http://creativecommons.org/licenses/by/4.0/

In the context of an increasing popularity of data-driven models to represent dynamical systems, many machine learning-based implementations of the Koopman operator have recently been proposed. However, the vast majority of those works are limited to deterministic predictions, while the knowledge of uncertainty is critical in fields like meteorology and climatology. In this work, we investigate the training of ensembles of models to produce stochastic outputs. We show through experiments on real remote sensing image time series that ensembles of independently trained models are highly overconfident and that using a training criterion that explicitly encourages the members to produce predictions with high inter-model variances greatly improves the uncertainty quantification of the ensembles.

[ { "created": "Mon, 11 Mar 2024 14:29:56 GMT", "version": "v1" }, { "created": "Wed, 13 Mar 2024 13:57:42 GMT", "version": "v2" } ]

2024-03-14

[ [ "Frion", "Anthony", "" ], [ "Drumetz", "Lucas", "" ], [ "Tochon", "Guillaume", "" ], [ "Mura", "Mauro Dalla", "" ], [ "Bey", "Albdeldjalil Aïssa El", "" ] ]

In the context of an increasing popularity of data-driven models to represent dynamical systems, many machine learning-based implementations of the Koopman operator have recently been proposed. However, the vast majority of those works are limited to deterministic predictions, while the knowledge of uncertainty is critical in fields like meteorology and climatology. In this work, we investigate the training of ensembles of models to produce stochastic outputs. We show through experiments on real remote sensing image time series that ensembles of independently trained models are highly overconfident and that using a training criterion that explicitly encourages the members to produce predictions with high inter-model variances greatly improves the uncertainty quantification of the ensembles.

1702.06968

Arthur-Jozsef Molnar

A Heuristic Process for GUI Widget Matching Across Application Versions

null

Annales Universitatis Scientiarum Budapest, Sectio Computatorica Vol 36, pages 255 - 275 (2012)

null

cs.SE

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper introduces an automated heuristic process able to achieve high accuracy when matching graphical user interface widgets across multiple versions of a target application. The proposed implementation is flexible as it allows full customization of the process and easy integration with existing tools for long term graphical user interface test case maintenance, software visualization and analysis.

[ { "created": "Wed, 22 Feb 2017 19:08:11 GMT", "version": "v1" } ]

2017-02-24

[ [ "Molnar", "Arthur-Jozsef", "" ] ]

This paper introduces an automated heuristic process able to achieve high accuracy when matching graphical user interface widgets across multiple versions of a target application. The proposed implementation is flexible as it allows full customization of the process and easy integration with existing tools for long term graphical user interface test case maintenance, software visualization and analysis.

2307.10936

Raphael Boige

Raphael Boige and Yannis Flet-Berliac and Arthur Flajolet and Guillaume Richard and Thomas Pierrot

PASTA: Pretrained Action-State Transformer Agents

null

cs.AI cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Self-supervised learning has brought about a revolutionary paradigm shift in various computing domains, including NLP, vision, and biology. Recent approaches involve pre-training transformer models on vast amounts of unlabeled data, serving as a starting point for efficiently solving downstream tasks. In reinforcement learning, researchers have recently adapted these approaches, developing models pre-trained on expert trajectories. This advancement enables the models to tackle a broad spectrum of tasks, ranging from robotics to recommendation systems. However, existing methods mostly rely on intricate pre-training objectives tailored to specific downstream applications. This paper conducts a comprehensive investigation of models, referred to as pre-trained action-state transformer agents (PASTA). Our study covers a unified methodology and covers an extensive set of general downstream tasks including behavioral cloning, offline RL, sensor failure robustness, and dynamics change adaptation. Our objective is to systematically compare various design choices and offer valuable insights that will aid practitioners in developing robust models. Key highlights of our study include tokenization at the component level for actions and states, the use of fundamental pre-training objectives such as next token prediction or masked language modeling, simultaneous training of models across multiple domains, and the application of various fine-tuning strategies. In this study, the developed models contain fewer than 7 million parameters allowing a broad community to use these models and reproduce our experiments. We hope that this study will encourage further research into the use of transformers with first principle design choices to represent RL trajectories and contribute to robust policy learning.

[ { "created": "Thu, 20 Jul 2023 15:09:06 GMT", "version": "v1" }, { "created": "Mon, 4 Dec 2023 10:15:26 GMT", "version": "v2" } ]

2023-12-05

[ [ "Boige", "Raphael", "" ], [ "Flet-Berliac", "Yannis", "" ], [ "Flajolet", "Arthur", "" ], [ "Richard", "Guillaume", "" ], [ "Pierrot", "Thomas", "" ] ]

Self-supervised learning has brought about a revolutionary paradigm shift in various computing domains, including NLP, vision, and biology. Recent approaches involve pre-training transformer models on vast amounts of unlabeled data, serving as a starting point for efficiently solving downstream tasks. In reinforcement learning, researchers have recently adapted these approaches, developing models pre-trained on expert trajectories. This advancement enables the models to tackle a broad spectrum of tasks, ranging from robotics to recommendation systems. However, existing methods mostly rely on intricate pre-training objectives tailored to specific downstream applications. This paper conducts a comprehensive investigation of models, referred to as pre-trained action-state transformer agents (PASTA). Our study covers a unified methodology and covers an extensive set of general downstream tasks including behavioral cloning, offline RL, sensor failure robustness, and dynamics change adaptation. Our objective is to systematically compare various design choices and offer valuable insights that will aid practitioners in developing robust models. Key highlights of our study include tokenization at the component level for actions and states, the use of fundamental pre-training objectives such as next token prediction or masked language modeling, simultaneous training of models across multiple domains, and the application of various fine-tuning strategies. In this study, the developed models contain fewer than 7 million parameters allowing a broad community to use these models and reproduce our experiments. We hope that this study will encourage further research into the use of transformers with first principle design choices to represent RL trajectories and contribute to robust policy learning.

1612.00604

Andrii Maksai

Andrii Maksai, Xinchao Wang, Francois Fleuret, and Pascal Fua

Globally Consistent Multi-People Tracking using Motion Patterns

8 pages, 7 figures. 11 pages supplementary

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Many state-of-the-art approaches to people tracking rely on detecting them in each frame independently, grouping detections into short but reliable trajectory segments, and then further grouping them into full trajectories. This grouping typically relies on imposing local smoothness constraints but almost never on enforcing more global constraints on the trajectories. In this paper, we propose an approach to imposing global consistency by first inferring behavioral patterns from the ground truth and then using them to guide the tracking algorithm. When used in conjunction with several state-of-the-art algorithms, this further increases their already good performance. Furthermore, we propose an unsupervised scheme that yields almost similar improvements without the need for ground truth.

[ { "created": "Fri, 2 Dec 2016 09:24:30 GMT", "version": "v1" } ]

2016-12-05

[ [ "Maksai", "Andrii", "" ], [ "Wang", "Xinchao", "" ], [ "Fleuret", "Francois", "" ], [ "Fua", "Pascal", "" ] ]

Many state-of-the-art approaches to people tracking rely on detecting them in each frame independently, grouping detections into short but reliable trajectory segments, and then further grouping them into full trajectories. This grouping typically relies on imposing local smoothness constraints but almost never on enforcing more global constraints on the trajectories. In this paper, we propose an approach to imposing global consistency by first inferring behavioral patterns from the ground truth and then using them to guide the tracking algorithm. When used in conjunction with several state-of-the-art algorithms, this further increases their already good performance. Furthermore, we propose an unsupervised scheme that yields almost similar improvements without the need for ground truth.

2311.11532

Gustavo Silva

Gustavo Silva, Paul Rodriguez

Optimal Hyperparameter $\epsilon$ for Adaptive Stochastic Optimizers through Gradient Histograms

null

cs.LG cs.AI

http://creativecommons.org/licenses/by-nc-nd/4.0/

Optimizers are essential components for successfully training deep neural network models. In order to achieve the best performance from such models, designers need to carefully choose the optimizer hyperparameters. However, this can be a computationally expensive and time-consuming process. Although it is known that all optimizer hyperparameters must be tuned for maximum performance, there is still a lack of clarity regarding the individual influence of minor priority hyperparameters, including the safeguard factor $\epsilon$ and momentum factor $\beta$, in leading adaptive optimizers (specifically, those based on the Adam optimizers). In this manuscript, we introduce a new framework based on gradient histograms to analyze and justify important attributes of adaptive optimizers, such as their optimal performance and the relationships and dependencies among hyperparameters. Furthermore, we propose a novel gradient histogram-based algorithm that automatically estimates a reduced and accurate search space for the safeguard hyperparameter $\epsilon$, where the optimal value can be easily found.

[ { "created": "Mon, 20 Nov 2023 04:34:19 GMT", "version": "v1" } ]

2023-11-21

[ [ "Silva", "Gustavo", "" ], [ "Rodriguez", "Paul", "" ] ]

Optimizers are essential components for successfully training deep neural network models. In order to achieve the best performance from such models, designers need to carefully choose the optimizer hyperparameters. However, this can be a computationally expensive and time-consuming process. Although it is known that all optimizer hyperparameters must be tuned for maximum performance, there is still a lack of clarity regarding the individual influence of minor priority hyperparameters, including the safeguard factor $\epsilon$ and momentum factor $\beta$, in leading adaptive optimizers (specifically, those based on the Adam optimizers). In this manuscript, we introduce a new framework based on gradient histograms to analyze and justify important attributes of adaptive optimizers, such as their optimal performance and the relationships and dependencies among hyperparameters. Furthermore, we propose a novel gradient histogram-based algorithm that automatically estimates a reduced and accurate search space for the safeguard hyperparameter $\epsilon$, where the optimal value can be easily found.

2302.02187

Niklas K\"uhl Prof Dr

Max Schemmer, Niklas K\"uhl, Carina Benz, Andrea Bartos, Gerhard Satzger

Appropriate Reliance on AI Advice: Conceptualization and the Effect of Explanations

arXiv admin note: text overlap with arXiv:2204.06916

ACM 28th International Conference on Intelligent User Interfaces (IUI), 2023

10.1145/3581641.3584066

null

cs.AI cs.HC

http://creativecommons.org/licenses/by-nc-nd/4.0/

AI advice is becoming increasingly popular, e.g., in investment and medical treatment decisions. As this advice is typically imperfect, decision-makers have to exert discretion as to whether actually follow that advice: they have to "appropriately" rely on correct and turn down incorrect advice. However, current research on appropriate reliance still lacks a common definition as well as an operational measurement concept. Additionally, no in-depth behavioral experiments have been conducted that help understand the factors influencing this behavior. In this paper, we propose Appropriateness of Reliance (AoR) as an underlying, quantifiable two-dimensional measurement concept. We develop a research model that analyzes the effect of providing explanations for AI advice. In an experiment with 200 participants, we demonstrate how these explanations influence the AoR, and, thus, the effectiveness of AI advice. Our work contributes fundamental concepts for the analysis of reliance behavior and the purposeful design of AI advisors.

[ { "created": "Sat, 4 Feb 2023 15:48:24 GMT", "version": "v1" }, { "created": "Tue, 7 Feb 2023 07:47:43 GMT", "version": "v2" }, { "created": "Thu, 13 Apr 2023 08:50:16 GMT", "version": "v3" } ]

2023-04-14

[ [ "Schemmer", "Max", "" ], [ "Kühl", "Niklas", "" ], [ "Benz", "Carina", "" ], [ "Bartos", "Andrea", "" ], [ "Satzger", "Gerhard", "" ] ]

AI advice is becoming increasingly popular, e.g., in investment and medical treatment decisions. As this advice is typically imperfect, decision-makers have to exert discretion as to whether actually follow that advice: they have to "appropriately" rely on correct and turn down incorrect advice. However, current research on appropriate reliance still lacks a common definition as well as an operational measurement concept. Additionally, no in-depth behavioral experiments have been conducted that help understand the factors influencing this behavior. In this paper, we propose Appropriateness of Reliance (AoR) as an underlying, quantifiable two-dimensional measurement concept. We develop a research model that analyzes the effect of providing explanations for AI advice. In an experiment with 200 participants, we demonstrate how these explanations influence the AoR, and, thus, the effectiveness of AI advice. Our work contributes fundamental concepts for the analysis of reliance behavior and the purposeful design of AI advisors.

1101.1815

Suvansh Lal Mr.

Suvansh Lal, Mohit Jain, Vikrant Chaplot

Approaches to Formal Verification of Security Protocols

null

cs.CR

http://creativecommons.org/licenses/by/3.0/

In recent times, many protocols have been proposed to provide security for various information and communication systems. Such protocols must be tested for their functional correctness before they are used in practice. Application of formal methods for verification of security protocols would enhance their reliability thereby, increasing the usability of systems that employ them. Thus, formal verification of security protocols has become a key issue in computer and communications security. In this paper we present, analyze and compare some prevalent approaches towards verification of secure systems. We follow the notion of - same goal through different approaches - as we formally analyze the Needham Schroeder Public Key protocol for Lowe's attack using each of our presented approaches.

[ { "created": "Mon, 10 Jan 2011 13:53:25 GMT", "version": "v1" } ]

2011-01-11

[ [ "Lal", "Suvansh", "" ], [ "Jain", "Mohit", "" ], [ "Chaplot", "Vikrant", "" ] ]

In recent times, many protocols have been proposed to provide security for various information and communication systems. Such protocols must be tested for their functional correctness before they are used in practice. Application of formal methods for verification of security protocols would enhance their reliability thereby, increasing the usability of systems that employ them. Thus, formal verification of security protocols has become a key issue in computer and communications security. In this paper we present, analyze and compare some prevalent approaches towards verification of secure systems. We follow the notion of - same goal through different approaches - as we formally analyze the Needham Schroeder Public Key protocol for Lowe's attack using each of our presented approaches.

2405.11530

Sejik Park

Learning More Generalized Experts by Merging Experts in Mixture-of-Experts

12 pages, 3 figures

null

cs.LG

http://creativecommons.org/licenses/by/4.0/

We observe that incorporating a shared layer in a mixture-of-experts can lead to performance degradation. This leads us to hypothesize that learning shared features poses challenges in deep learning, potentially caused by the same feature being learned as various different features. To address this issue, we track each expert's usage frequency and merge the two most frequently selected experts. We then update the least frequently selected expert using the combination of experts. This approach, combined with the subsequent learning of the router's expert selection, allows the model to determine if the most frequently selected experts have learned the same feature differently. If they have, the combined expert can be further trained to learn a more general feature. Consequently, our algorithm enhances transfer learning and mitigates catastrophic forgetting when applied to multi-domain task incremental learning.

[ { "created": "Sun, 19 May 2024 11:55:48 GMT", "version": "v1" } ]

2024-05-21

[ [ "Park", "Sejik", "" ] ]

We observe that incorporating a shared layer in a mixture-of-experts can lead to performance degradation. This leads us to hypothesize that learning shared features poses challenges in deep learning, potentially caused by the same feature being learned as various different features. To address this issue, we track each expert's usage frequency and merge the two most frequently selected experts. We then update the least frequently selected expert using the combination of experts. This approach, combined with the subsequent learning of the router's expert selection, allows the model to determine if the most frequently selected experts have learned the same feature differently. If they have, the combined expert can be further trained to learn a more general feature. Consequently, our algorithm enhances transfer learning and mitigates catastrophic forgetting when applied to multi-domain task incremental learning.

1701.01216

Tony T. Luo

T. Luo, S. S. Kanhere, H-P. Tan, F. Wu, and H. Wu

Crowdsourcing with Tullock contests: A new perspective

9 pages, 4 figures, 3 tables

Proc. IEEE INFOCOM, 2015, pp. 2515-2523

10.1109/INFOCOM.2015.7218641

null

cs.GT cs.HC cs.MA cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Incentive mechanisms for crowdsourcing have been extensively studied under the framework of all-pay auctions. Along a distinct line, this paper proposes to use Tullock contests as an alternative tool to design incentive mechanisms for crowdsourcing. We are inspired by the conduciveness of Tullock contests to attracting user entry (yet not necessarily a higher revenue) in other domains. In this paper, we explore a new dimension in optimal Tullock contest design, by superseding the contest prize---which is fixed in conventional Tullock contests---with a prize function that is dependent on the (unknown) winner's contribution, in order to maximize the crowdsourcer's utility. We show that this approach leads to attractive practical advantages: (a) it is well-suited for rapid prototyping in fully distributed web agents and smartphone apps; (b) it overcomes the disincentive to participate caused by players' antagonism to an increasing number of rivals. Furthermore, we optimize conventional, fixed-prize Tullock contests to construct the most superior benchmark to compare against our mechanism. Through extensive evaluations, we show that our mechanism significantly outperforms the optimal benchmark, by over three folds on the crowdsourcer's utility cum profit and up to nine folds on the players' social welfare.

[ { "created": "Thu, 5 Jan 2017 05:44:25 GMT", "version": "v1" } ]

2017-01-06

[ [ "Luo", "T.", "" ], [ "Kanhere", "S. S.", "" ], [ "Tan", "H-P.", "" ], [ "Wu", "F.", "" ], [ "Wu", "H.", "" ] ]

Incentive mechanisms for crowdsourcing have been extensively studied under the framework of all-pay auctions. Along a distinct line, this paper proposes to use Tullock contests as an alternative tool to design incentive mechanisms for crowdsourcing. We are inspired by the conduciveness of Tullock contests to attracting user entry (yet not necessarily a higher revenue) in other domains. In this paper, we explore a new dimension in optimal Tullock contest design, by superseding the contest prize---which is fixed in conventional Tullock contests---with a prize function that is dependent on the (unknown) winner's contribution, in order to maximize the crowdsourcer's utility. We show that this approach leads to attractive practical advantages: (a) it is well-suited for rapid prototyping in fully distributed web agents and smartphone apps; (b) it overcomes the disincentive to participate caused by players' antagonism to an increasing number of rivals. Furthermore, we optimize conventional, fixed-prize Tullock contests to construct the most superior benchmark to compare against our mechanism. Through extensive evaluations, we show that our mechanism significantly outperforms the optimal benchmark, by over three folds on the crowdsourcer's utility cum profit and up to nine folds on the players' social welfare.

1704.00355

Neha Gupta

Moses Charikar, Neha Gupta, Roy Schwartz

Local Guarantees in Graph Cuts and Clustering

null

cs.DS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Correlation Clustering is an elegant model that captures fundamental graph cut problems such as Min $s-t$ Cut, Multiway Cut, and Multicut, extensively studied in combinatorial optimization. Here, we are given a graph with edges labeled $+$ or $-$ and the goal is to produce a clustering that agrees with the labels as much as possible: $+$ edges within clusters and $-$ edges across clusters. The classical approach towards Correlation Clustering (and other graph cut problems) is to optimize a global objective. We depart from this and study local objectives: minimizing the maximum number of disagreements for edges incident on a single node, and the analogous max min agreements objective. This naturally gives rise to a family of basic min-max graph cut problems. A prototypical representative is Min Max $s-t$ Cut: find an $s-t$ cut minimizing the largest number of cut edges incident on any node. We present the following results: $(1)$ an $O(\sqrt{n})$-approximation for the problem of minimizing the maximum total weight of disagreement edges incident on any node (thus providing the first known approximation for the above family of min-max graph cut problems), $(2)$ a remarkably simple $7$-approximation for minimizing local disagreements in complete graphs (improving upon the previous best known approximation of $48$), and $(3)$ a $1/(2+\varepsilon)$-approximation for maximizing the minimum total weight of agreement edges incident on any node, hence improving upon the $1/(4+\varepsilon)$-approximation that follows from the study of approximate pure Nash equilibria in cut and party affiliation games.

[ { "created": "Sun, 2 Apr 2017 19:34:22 GMT", "version": "v1" } ]

2017-04-04

[ [ "Charikar", "Moses", "" ], [ "Gupta", "Neha", "" ], [ "Schwartz", "Roy", "" ] ]

Correlation Clustering is an elegant model that captures fundamental graph cut problems such as Min $s-t$ Cut, Multiway Cut, and Multicut, extensively studied in combinatorial optimization. Here, we are given a graph with edges labeled $+$ or $-$ and the goal is to produce a clustering that agrees with the labels as much as possible: $+$ edges within clusters and $-$ edges across clusters. The classical approach towards Correlation Clustering (and other graph cut problems) is to optimize a global objective. We depart from this and study local objectives: minimizing the maximum number of disagreements for edges incident on a single node, and the analogous max min agreements objective. This naturally gives rise to a family of basic min-max graph cut problems. A prototypical representative is Min Max $s-t$ Cut: find an $s-t$ cut minimizing the largest number of cut edges incident on any node. We present the following results: $(1)$ an $O(\sqrt{n})$-approximation for the problem of minimizing the maximum total weight of disagreement edges incident on any node (thus providing the first known approximation for the above family of min-max graph cut problems), $(2)$ a remarkably simple $7$-approximation for minimizing local disagreements in complete graphs (improving upon the previous best known approximation of $48$), and $(3)$ a $1/(2+\varepsilon)$-approximation for maximizing the minimum total weight of agreement edges incident on any node, hence improving upon the $1/(4+\varepsilon)$-approximation that follows from the study of approximate pure Nash equilibria in cut and party affiliation games.

1502.03358

Farshad Naghibi

Farshad Naghibi, Somayeh Salimi, Mikael Skoglund

The CEO Problem with Secrecy Constraints

Accepted for publication in IEEE Transactions on Information Forensics and Security, 17 pages, 4 figures

null

10.1109/TIFS.2015.2404134

null

cs.IT math.IT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We study a lossy source coding problem with secrecy constraints in which a remote information source should be transmitted to a single destination via multiple agents in the presence of a passive eavesdropper. The agents observe noisy versions of the source and independently encode and transmit their observations to the destination via noiseless rate-limited links. The destination should estimate the remote source based on the information received from the agents within a certain mean distortion threshold. The eavesdropper, with access to side information correlated to the source, is able to listen in on one of the links from the agents to the destination in order to obtain as much information as possible about the source. This problem can be viewed as the so-called CEO problem with additional secrecy constraints. We establish inner and outer bounds on the rate-distortion-equivocation region of this problem. We also obtain the region in special cases where the bounds are tight. Furthermore, we study the quadratic Gaussian case and provide the optimal rate-distortion-equivocation region when the eavesdropper has no side information and an achievable region for a more general setup with side information at the eavesdropper.

[ { "created": "Wed, 11 Feb 2015 16:19:21 GMT", "version": "v1" } ]

2015-02-19

[ [ "Naghibi", "Farshad", "" ], [ "Salimi", "Somayeh", "" ], [ "Skoglund", "Mikael", "" ] ]

We study a lossy source coding problem with secrecy constraints in which a remote information source should be transmitted to a single destination via multiple agents in the presence of a passive eavesdropper. The agents observe noisy versions of the source and independently encode and transmit their observations to the destination via noiseless rate-limited links. The destination should estimate the remote source based on the information received from the agents within a certain mean distortion threshold. The eavesdropper, with access to side information correlated to the source, is able to listen in on one of the links from the agents to the destination in order to obtain as much information as possible about the source. This problem can be viewed as the so-called CEO problem with additional secrecy constraints. We establish inner and outer bounds on the rate-distortion-equivocation region of this problem. We also obtain the region in special cases where the bounds are tight. Furthermore, we study the quadratic Gaussian case and provide the optimal rate-distortion-equivocation region when the eavesdropper has no side information and an achievable region for a more general setup with side information at the eavesdropper.

2007.00236

Cunxiang Wang

Cunxiang Wang, Shuailong Liang, Yili Jin, Yilong Wang, Xiaodan Zhu and Yue Zhang

SemEval-2020 Task 4: Commonsense Validation and Explanation

Task description paper of SemEval-2020 Task 4: Commonsense Validation and Explanation

null

cs.CL cs.AI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this paper, we present SemEval-2020 Task 4, Commonsense Validation and Explanation (ComVE), which includes three subtasks, aiming to evaluate whether a system can distinguish a natural language statement that makes sense to humans from one that does not, and provide the reasons. Specifically, in our first subtask, the participating systems are required to choose from two natural language statements of similar wording the one that makes sense and the one does not. The second subtask additionally asks a system to select the key reason from three options why a given statement does not make sense. In the third subtask, a participating system needs to generate the reason. We finally attracted 39 teams participating at least one of the three subtasks. For Subtask A and Subtask B, the performances of top-ranked systems are close to that of humans. However, for Subtask C, there is still a relatively large gap between systems and human performance. The dataset used in our task can be found at https://github.com/wangcunxiang/SemEval2020- Task4-Commonsense-Validation-and-Explanation; The leaderboard can be found at https://competitions.codalab.org/competitions/21080#results.

[ { "created": "Wed, 1 Jul 2020 04:41:05 GMT", "version": "v1" }, { "created": "Mon, 3 Aug 2020 15:13:40 GMT", "version": "v2" } ]

2020-08-04

[ [ "Wang", "Cunxiang", "" ], [ "Liang", "Shuailong", "" ], [ "Jin", "Yili", "" ], [ "Wang", "Yilong", "" ], [ "Zhu", "Xiaodan", "" ], [ "Zhang", "Yue", "" ] ]

In this paper, we present SemEval-2020 Task 4, Commonsense Validation and Explanation (ComVE), which includes three subtasks, aiming to evaluate whether a system can distinguish a natural language statement that makes sense to humans from one that does not, and provide the reasons. Specifically, in our first subtask, the participating systems are required to choose from two natural language statements of similar wording the one that makes sense and the one does not. The second subtask additionally asks a system to select the key reason from three options why a given statement does not make sense. In the third subtask, a participating system needs to generate the reason. We finally attracted 39 teams participating at least one of the three subtasks. For Subtask A and Subtask B, the performances of top-ranked systems are close to that of humans. However, for Subtask C, there is still a relatively large gap between systems and human performance. The dataset used in our task can be found at https://github.com/wangcunxiang/SemEval2020- Task4-Commonsense-Validation-and-Explanation; The leaderboard can be found at https://competitions.codalab.org/competitions/21080#results.

2304.13518

Yuqi Han

Yuqi Han and Tao Yu and Xiaohang Yu and Yuwang Wang and Qionghai Dai

Super-NeRF: View-consistent Detail Generation for NeRF super-resolution

null

cs.CV

http://creativecommons.org/licenses/by/4.0/

The neural radiance field (NeRF) achieved remarkable success in modeling 3D scenes and synthesizing high-fidelity novel views. However, existing NeRF-based methods focus more on the make full use of the image resolution to generate novel views, but less considering the generation of details under the limited input resolution. In analogy to the extensive usage of image super-resolution, NeRF super-resolution is an effective way to generate the high-resolution implicit representation of 3D scenes and holds great potential applications. Up to now, such an important topic is still under-explored. In this paper, we propose a NeRF super-resolution method, named Super-NeRF, to generate high-resolution NeRF from only low-resolution inputs. Given multi-view low-resolution images, Super-NeRF constructs a consistency-controlling super-resolution module to generate view-consistent high-resolution details for NeRF. Specifically, an optimizable latent code is introduced for each low-resolution input image to control the 2D super-resolution images to converge to the view-consistent output. The latent codes of each low-resolution image are optimized synergistically with the target Super-NeRF representation to fully utilize the view consistency constraint inherent in NeRF construction. We verify the effectiveness of Super-NeRF on synthetic, real-world, and AI-generated NeRF datasets. Super-NeRF achieves state-of-the-art NeRF super-resolution performance on high-resolution detail generation and cross-view consistency.

[ { "created": "Wed, 26 Apr 2023 12:54:40 GMT", "version": "v1" } ]

2023-04-27

[ [ "Han", "Yuqi", "" ], [ "Yu", "Tao", "" ], [ "Yu", "Xiaohang", "" ], [ "Wang", "Yuwang", "" ], [ "Dai", "Qionghai", "" ] ]

The neural radiance field (NeRF) achieved remarkable success in modeling 3D scenes and synthesizing high-fidelity novel views. However, existing NeRF-based methods focus more on the make full use of the image resolution to generate novel views, but less considering the generation of details under the limited input resolution. In analogy to the extensive usage of image super-resolution, NeRF super-resolution is an effective way to generate the high-resolution implicit representation of 3D scenes and holds great potential applications. Up to now, such an important topic is still under-explored. In this paper, we propose a NeRF super-resolution method, named Super-NeRF, to generate high-resolution NeRF from only low-resolution inputs. Given multi-view low-resolution images, Super-NeRF constructs a consistency-controlling super-resolution module to generate view-consistent high-resolution details for NeRF. Specifically, an optimizable latent code is introduced for each low-resolution input image to control the 2D super-resolution images to converge to the view-consistent output. The latent codes of each low-resolution image are optimized synergistically with the target Super-NeRF representation to fully utilize the view consistency constraint inherent in NeRF construction. We verify the effectiveness of Super-NeRF on synthetic, real-world, and AI-generated NeRF datasets. Super-NeRF achieves state-of-the-art NeRF super-resolution performance on high-resolution detail generation and cross-view consistency.

1809.05606

Yimin Yang

Yimin Yang, Q.M.Jonathan Wu, Xiexing Feng, Thangarajah Akilan

Non-iterative recomputation of dense layers for performance improvement of DCNN

11

null

cs.LG stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

An iterative method of learning has become a paradigm for training deep convolutional neural networks (DCNN). However, utilizing a non-iterative learning strategy can accelerate the training process of the DCNN and surprisingly such approach has been rarely explored by the deep learning (DL) community. It motivates this paper to introduce a non-iterative learning strategy that eliminates the backpropagation (BP) at the top dense or fully connected (FC) layers of DCNN, resulting in, lower training time and higher performance. The proposed method exploits the Moore-Penrose Inverse to pull back the current residual error to each FC layer, generating well-generalized features. Then using the recomputed features, i.e., the new generalized features the weights of each FC layer is computed according to the Moore-Penrose Inverse. We evaluate the proposed approach on six widely accepted object recognition benchmark datasets: Scene-15, CIFAR-10, CIFAR-100, SUN-397, Places365, and ImageNet. The experimental results show that the proposed method obtains significant improvements over 30 state-of-the-art methods. Interestingly, it also indicates that any DCNN with the proposed method can provide better performance than the same network with its original training based on BP.

[ { "created": "Fri, 14 Sep 2018 22:24:52 GMT", "version": "v1" } ]

2018-09-18

[ [ "Yang", "Yimin", "" ], [ "Wu", "Q. M. Jonathan", "" ], [ "Feng", "Xiexing", "" ], [ "Akilan", "Thangarajah", "" ] ]

An iterative method of learning has become a paradigm for training deep convolutional neural networks (DCNN). However, utilizing a non-iterative learning strategy can accelerate the training process of the DCNN and surprisingly such approach has been rarely explored by the deep learning (DL) community. It motivates this paper to introduce a non-iterative learning strategy that eliminates the backpropagation (BP) at the top dense or fully connected (FC) layers of DCNN, resulting in, lower training time and higher performance. The proposed method exploits the Moore-Penrose Inverse to pull back the current residual error to each FC layer, generating well-generalized features. Then using the recomputed features, i.e., the new generalized features the weights of each FC layer is computed according to the Moore-Penrose Inverse. We evaluate the proposed approach on six widely accepted object recognition benchmark datasets: Scene-15, CIFAR-10, CIFAR-100, SUN-397, Places365, and ImageNet. The experimental results show that the proposed method obtains significant improvements over 30 state-of-the-art methods. Interestingly, it also indicates that any DCNN with the proposed method can provide better performance than the same network with its original training based on BP.

1309.0871

EPTCS

Oded Maler (CNRS-VERIMAG, University of Grenoble), \'Ad\'am M. Hal\'asz (Department of Methematics, West Virginia University), Olivier Lebeltel (CNRS-VERIMAG, University of Grenoble), Ouri Maler (Grenoble)

Exploring the Dynamics of Mass Action Systems

In Proceedings HSB 2013, arXiv:1308.5724

EPTCS 125, 2013, pp. 84-91

10.4204/EPTCS.125.6

null

cs.CE cs.CY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We present the Populus toolkit for exploring the dynamics of mass action systems under different assumptions.

[ { "created": "Tue, 3 Sep 2013 23:41:36 GMT", "version": "v1" } ]

2013-09-05

[ [ "Maler", "Oded", "", "CNRS-VERIMAG, University of Grenoble" ], [ "Halász", "Ádám M.", "", "Department of Methematics, West Virginia University" ], [ "Lebeltel", "Olivier", "", "CNRS-VERIMAG, University of Grenoble" ], [ "Maler", "Ouri", "", "Grenoble" ] ]

We present the Populus toolkit for exploring the dynamics of mass action systems under different assumptions.

2403.05193

Martina Benini

Martina Benini, Silvia Gallucci, Marta Bonato, Marta Parazzini, Gabriella Tognola

Evaluation of Road User Radio-Frequency Exposure Levels in an Urban Environment from Vehicular Antennas and the Infrastructure in ITS-G5 5.9 GHz Communication

null

cs.NI

http://creativecommons.org/licenses/by/4.0/

This study aims to investigate the variability of exposure levels among road users generated in a realistic urban scenario by Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies operating at 5.9 GHz. The exposure levels were evaluated in terms of whole-body Specific Absorption Rate (wbSAR) [W/kg] in three different human models, ranging from children to adults. We calculated the electromagnetic field exposure level generated by V2V and V2I using raytracing and we assessed wbSAR resulting in urban exposure scenarios with an increasing number of transmitting antennas. Whole-body SAR was generally very low, on the order of 10^-4 W/kg. The maximum wbSAR, of 4.9x10^-4 W/kg, was obtained in the worst-case exposure condition comprising more than one transmitting vehicle and was found in the adult model for a distance within 10 m from the transmitting cars. We found that the height of the human model highly impacted the exposure level. Namely, the child (which is the shortest human model) was generally much less exposed than adults. All the wbSAR values found by varying the number of transmitting antennas, the distance of the road user from the antennas, and the type of human model (adult vs. child) were very well below the limits set by the ICNIRP and IEEE guidelines of 0.08 W/kg for human exposure in the 100 kHz - 300 GHz range.

[ { "created": "Fri, 8 Mar 2024 10:14:40 GMT", "version": "v1" } ]

2024-03-11

[ [ "Benini", "Martina", "" ], [ "Gallucci", "Silvia", "" ], [ "Bonato", "Marta", "" ], [ "Parazzini", "Marta", "" ], [ "Tognola", "Gabriella", "" ] ]

This study aims to investigate the variability of exposure levels among road users generated in a realistic urban scenario by Vehicle-to-Vehicle (V2V) and Vehicle-to-Infrastructure (V2I) communication technologies operating at 5.9 GHz. The exposure levels were evaluated in terms of whole-body Specific Absorption Rate (wbSAR) [W/kg] in three different human models, ranging from children to adults. We calculated the electromagnetic field exposure level generated by V2V and V2I using raytracing and we assessed wbSAR resulting in urban exposure scenarios with an increasing number of transmitting antennas. Whole-body SAR was generally very low, on the order of 10^-4 W/kg. The maximum wbSAR, of 4.9x10^-4 W/kg, was obtained in the worst-case exposure condition comprising more than one transmitting vehicle and was found in the adult model for a distance within 10 m from the transmitting cars. We found that the height of the human model highly impacted the exposure level. Namely, the child (which is the shortest human model) was generally much less exposed than adults. All the wbSAR values found by varying the number of transmitting antennas, the distance of the road user from the antennas, and the type of human model (adult vs. child) were very well below the limits set by the ICNIRP and IEEE guidelines of 0.08 W/kg for human exposure in the 100 kHz - 300 GHz range.

2408.03945

Kristina Schaaff

Kristina Schaaff and Marc-Andr\'e Heidelmann

Impacts of Anthropomorphizing Large Language Models in Learning Environments

Presented at Affective Computing Pre-Conference at ISRE 2024

null

cs.CL cs.AI cs.CY cs.HC

http://creativecommons.org/licenses/by/4.0/

Large Language Models (LLMs) are increasingly being used in learning environments to support teaching-be it as learning companions or as tutors. With our contribution, we aim to discuss the implications of the anthropomorphization of LLMs in learning environments on educational theory to build a foundation for more effective learning outcomes and understand their emotional impact on learners. According to the media equation, people tend to respond to media in the same way as they would respond to another person. A study conducted by the Georgia Institute of Technology showed that chatbots can be successfully implemented in learning environments. In this study, learners in selected online courses were unable to distinguish the chatbot from a "real" teacher. As LLM-based chatbots such as OpenAI's GPT series are increasingly used in educational tools, it is important to understand how the attribution processes to LLM-based chatbots in terms of anthropomorphization affect learners' emotions.

[ { "created": "Mon, 22 Jul 2024 06:28:54 GMT", "version": "v1" } ]

2024-08-09

[ [ "Schaaff", "Kristina", "" ], [ "Heidelmann", "Marc-André", "" ] ]

Large Language Models (LLMs) are increasingly being used in learning environments to support teaching-be it as learning companions or as tutors. With our contribution, we aim to discuss the implications of the anthropomorphization of LLMs in learning environments on educational theory to build a foundation for more effective learning outcomes and understand their emotional impact on learners. According to the media equation, people tend to respond to media in the same way as they would respond to another person. A study conducted by the Georgia Institute of Technology showed that chatbots can be successfully implemented in learning environments. In this study, learners in selected online courses were unable to distinguish the chatbot from a "real" teacher. As LLM-based chatbots such as OpenAI's GPT series are increasingly used in educational tools, it is important to understand how the attribution processes to LLM-based chatbots in terms of anthropomorphization affect learners' emotions.

2310.04197

Angel Casanova

\'Angel Casanova Bienzobas, Alfonso S\'anchez-Maci\'an

Threat Trekker: An Approach to Cyber Threat Hunting

I am disseminating this outcome to all of you, despite the fact that the results may appear somewhat idealistic, given that certain datasets utilized for the training of the machine learning model comprise simulated data

null

cs.CR

http://creativecommons.org/licenses/by-nc-nd/4.0/

Threat hunting is a proactive methodology for exploring, detecting and mitigating cyberattacks within complex environments. As opposed to conventional detection systems, threat hunting strategies assume adversaries have infiltrated the system; as a result they proactively search out any unusual patterns or activities which might indicate intrusion attempts. Historically, this endeavour has been pursued using three investigation methodologies: (1) Hypothesis-Driven Investigations; (2) Indicator of Compromise (IOC); and (3) High-level machine learning analysis-based approaches. Therefore, this paper introduces a novel machine learning paradigm known as Threat Trekker. This proposal utilizes connectors to feed data directly into an event streaming channel for processing by the algorithm and provide feedback back into its host network. Conclusions drawn from these experiments clearly establish the efficacy of employing machine learning for classifying more subtle attacks.

[ { "created": "Fri, 6 Oct 2023 12:29:41 GMT", "version": "v1" } ]

2023-10-09

[ [ "Bienzobas", "Ángel Casanova", "" ], [ "Sánchez-Macián", "Alfonso", "" ] ]

Threat hunting is a proactive methodology for exploring, detecting and mitigating cyberattacks within complex environments. As opposed to conventional detection systems, threat hunting strategies assume adversaries have infiltrated the system; as a result they proactively search out any unusual patterns or activities which might indicate intrusion attempts. Historically, this endeavour has been pursued using three investigation methodologies: (1) Hypothesis-Driven Investigations; (2) Indicator of Compromise (IOC); and (3) High-level machine learning analysis-based approaches. Therefore, this paper introduces a novel machine learning paradigm known as Threat Trekker. This proposal utilizes connectors to feed data directly into an event streaming channel for processing by the algorithm and provide feedback back into its host network. Conclusions drawn from these experiments clearly establish the efficacy of employing machine learning for classifying more subtle attacks.

2205.06255

Qianqian Wang

Qianqian Wang, Zhengqi Li, David Salesin, Noah Snavely, Brian Curless, Janne Kontkanen

3D Moments from Near-Duplicate Photos

CVPR 2022

null

cs.CV

http://creativecommons.org/licenses/by/4.0/

We introduce 3D Moments, a new computational photography effect. As input we take a pair of near-duplicate photos, i.e., photos of moving subjects from similar viewpoints, common in people's photo collections. As output, we produce a video that smoothly interpolates the scene motion from the first photo to the second, while also producing camera motion with parallax that gives a heightened sense of 3D. To achieve this effect, we represent the scene as a pair of feature-based layered depth images augmented with scene flow. This representation enables motion interpolation along with independent control of the camera viewpoint. Our system produces photorealistic space-time videos with motion parallax and scene dynamics, while plausibly recovering regions occluded in the original views. We conduct extensive experiments demonstrating superior performance over baselines on public datasets and in-the-wild photos. Project page: https://3d-moments.github.io/

[ { "created": "Thu, 12 May 2022 17:56:18 GMT", "version": "v1" } ]

2022-05-13

[ [ "Wang", "Qianqian", "" ], [ "Li", "Zhengqi", "" ], [ "Salesin", "David", "" ], [ "Snavely", "Noah", "" ], [ "Curless", "Brian", "" ], [ "Kontkanen", "Janne", "" ] ]

We introduce 3D Moments, a new computational photography effect. As input we take a pair of near-duplicate photos, i.e., photos of moving subjects from similar viewpoints, common in people's photo collections. As output, we produce a video that smoothly interpolates the scene motion from the first photo to the second, while also producing camera motion with parallax that gives a heightened sense of 3D. To achieve this effect, we represent the scene as a pair of feature-based layered depth images augmented with scene flow. This representation enables motion interpolation along with independent control of the camera viewpoint. Our system produces photorealistic space-time videos with motion parallax and scene dynamics, while plausibly recovering regions occluded in the original views. We conduct extensive experiments demonstrating superior performance over baselines on public datasets and in-the-wild photos. Project page: https://3d-moments.github.io/

2407.15373

Dizhi Ma

Dizhi Ma, Xiyun Hu, Jingyu Shi, Mayank Patel, Rahul Jain, Ziyi Liu, Zhengzhe Zhu and Karthik Ramani

avaTTAR: Table Tennis Stroke Training with On-body and Detached Visualization in Augmented Reality

null

cs.HC

http://creativecommons.org/licenses/by/4.0/

Table tennis stroke training is a critical aspect of player development. We designed a new augmented reality (AR) system, avaTTAR, for table tennis stroke training. The system provides both "on-body" (first-person view) and "detached" (third-person view) visual cues, enabling users to visualize target strokes and correct their attempts effectively with this dual perspectives setup. By employing a combination of pose estimation algorithms and IMU sensors, avaTTAR captures and reconstructs the 3D body pose and paddle orientation of users during practice, allowing real-time comparison with expert strokes. Through a user study, we affirm avaTTAR's capacity to amplify player experience and training results.

[ { "created": "Mon, 22 Jul 2024 04:47:16 GMT", "version": "v1" }, { "created": "Fri, 26 Jul 2024 15:13:46 GMT", "version": "v2" } ]

2024-07-29

[ [ "Ma", "Dizhi", "" ], [ "Hu", "Xiyun", "" ], [ "Shi", "Jingyu", "" ], [ "Patel", "Mayank", "" ], [ "Jain", "Rahul", "" ], [ "Liu", "Ziyi", "" ], [ "Zhu", "Zhengzhe", "" ], [ "Ramani", "Karthik", "" ] ]

Table tennis stroke training is a critical aspect of player development. We designed a new augmented reality (AR) system, avaTTAR, for table tennis stroke training. The system provides both "on-body" (first-person view) and "detached" (third-person view) visual cues, enabling users to visualize target strokes and correct their attempts effectively with this dual perspectives setup. By employing a combination of pose estimation algorithms and IMU sensors, avaTTAR captures and reconstructs the 3D body pose and paddle orientation of users during practice, allowing real-time comparison with expert strokes. Through a user study, we affirm avaTTAR's capacity to amplify player experience and training results.

2005.13102

Bangalore Ravi Kiran

Leonardo Gigli, B Ravi Kiran, Thomas Paul, Andres Serna, Nagarjuna Vemuri, Beatriz Marcotegui, Santiago Velasco-Forero

Road Segmentation on low resolution Lidar point clouds for autonomous vehicles

ISPRS 2020

null

cs.CV stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Point cloud datasets for perception tasks in the context of autonomous driving often rely on high resolution 64-layer Light Detection and Ranging (LIDAR) scanners. They are expensive to deploy on real-world autonomous driving sensor architectures which usually employ 16/32 layer LIDARs. We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. In our experiments the low resolution 16/32 layer LIDAR point clouds are simulated by subsampling the original 64 layer data, for subsequent transformation in to a feature map in the Bird-Eye-View (BEV) and SphericalView (SV) representations of the point cloud. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures. We demonstrate that this local normal feature in conjunction with classical features not only improves performance for binary road segmentation on full resolution point clouds, but it also reduces the negative impact on the accuracy when subsampling dense point clouds as compared to the usage of classical features alone. We assess our method with several experiments on two datasets: KITTI Road-segmentation benchmark and the recently released Semantic KITTI dataset.

[ { "created": "Wed, 27 May 2020 00:38:39 GMT", "version": "v1" } ]

2020-05-28

[ [ "Gigli", "Leonardo", "" ], [ "Kiran", "B Ravi", "" ], [ "Paul", "Thomas", "" ], [ "Serna", "Andres", "" ], [ "Vemuri", "Nagarjuna", "" ], [ "Marcotegui", "Beatriz", "" ], [ "Velasco-Forero", "Santiago", "" ] ]

Point cloud datasets for perception tasks in the context of autonomous driving often rely on high resolution 64-layer Light Detection and Ranging (LIDAR) scanners. They are expensive to deploy on real-world autonomous driving sensor architectures which usually employ 16/32 layer LIDARs. We evaluate the effect of subsampling image based representations of dense point clouds on the accuracy of the road segmentation task. In our experiments the low resolution 16/32 layer LIDAR point clouds are simulated by subsampling the original 64 layer data, for subsequent transformation in to a feature map in the Bird-Eye-View (BEV) and SphericalView (SV) representations of the point cloud. We introduce the usage of the local normal vector with the LIDAR's spherical coordinates as an input channel to existing LoDNN architectures. We demonstrate that this local normal feature in conjunction with classical features not only improves performance for binary road segmentation on full resolution point clouds, but it also reduces the negative impact on the accuracy when subsampling dense point clouds as compared to the usage of classical features alone. We assess our method with several experiments on two datasets: KITTI Road-segmentation benchmark and the recently released Semantic KITTI dataset.

2105.05026

Chongxuan Li

Guoqiang Wu, Chongxuan Li, Kun Xu, Jun Zhu

Rethinking and Reweighting the Univariate Losses for Multi-Label Ranking: Consistency and Generalization

null

cs.LG stat.ML

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

(Partial) ranking loss is a commonly used evaluation measure for multi-label classification, which is usually optimized with convex surrogates for computational efficiency. Prior theoretical work on multi-label ranking mainly focuses on (Fisher) consistency analyses. However, there is a gap between existing theory and practice -- some pairwise losses can lead to promising performance but lack consistency, while some univariate losses are consistent but usually have no clear superiority in practice. In this paper, we attempt to fill this gap through a systematic study from two complementary perspectives of consistency and generalization error bounds of learning algorithms. Our results show that learning algorithms with the consistent univariate loss have an error bound of $O(c)$ ($c$ is the number of labels), while algorithms with the inconsistent pairwise loss depend on $O(\sqrt{c})$ as shown in prior work. This explains that the latter can achieve better performance than the former in practice. Moreover, we present an inconsistent reweighted univariate loss-based learning algorithm that enjoys an error bound of $O(\sqrt{c})$ for promising performance as well as the computational efficiency of univariate losses. Finally, experimental results validate our theoretical analyses.

[ { "created": "Mon, 10 May 2021 09:23:27 GMT", "version": "v1" } ]

2021-05-12

[ [ "Wu", "Guoqiang", "" ], [ "Li", "Chongxuan", "" ], [ "Xu", "Kun", "" ], [ "Zhu", "Jun", "" ] ]

(Partial) ranking loss is a commonly used evaluation measure for multi-label classification, which is usually optimized with convex surrogates for computational efficiency. Prior theoretical work on multi-label ranking mainly focuses on (Fisher) consistency analyses. However, there is a gap between existing theory and practice -- some pairwise losses can lead to promising performance but lack consistency, while some univariate losses are consistent but usually have no clear superiority in practice. In this paper, we attempt to fill this gap through a systematic study from two complementary perspectives of consistency and generalization error bounds of learning algorithms. Our results show that learning algorithms with the consistent univariate loss have an error bound of $O(c)$ ($c$ is the number of labels), while algorithms with the inconsistent pairwise loss depend on $O(\sqrt{c})$ as shown in prior work. This explains that the latter can achieve better performance than the former in practice. Moreover, we present an inconsistent reweighted univariate loss-based learning algorithm that enjoys an error bound of $O(\sqrt{c})$ for promising performance as well as the computational efficiency of univariate losses. Finally, experimental results validate our theoretical analyses.

2403.14623

Zhicong Tang

Zhicong Tang, Tiankai Hang, Shuyang Gu, Dong Chen, Baining Guo

Simplified Diffusion Schr\"odinger Bridge

null

cs.LG cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

This paper introduces a novel theoretical simplification of the Diffusion Schr\"odinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance. By employing SGMs as an initial solution for DSB, our approach capitalizes on the strengths of both frameworks, ensuring a more efficient training process and improving the performance of SGM. We also propose a reparameterization technique that, despite theoretical approximations, practically improves the network's fitting capabilities. Our extensive experimental evaluations confirm the effectiveness of the simplified DSB, demonstrating its significant improvements. We believe the contributions of this work pave the way for advanced generative modeling.

[ { "created": "Thu, 21 Mar 2024 17:59:41 GMT", "version": "v1" }, { "created": "Wed, 27 Mar 2024 16:49:35 GMT", "version": "v2" }, { "created": "Mon, 27 May 2024 04:44:22 GMT", "version": "v3" }, { "created": "Tue, 13 Aug 2024 04:34:58 GMT", "version": "v4" } ]

2024-08-14

[ [ "Tang", "Zhicong", "" ], [ "Hang", "Tiankai", "" ], [ "Gu", "Shuyang", "" ], [ "Chen", "Dong", "" ], [ "Guo", "Baining", "" ] ]

This paper introduces a novel theoretical simplification of the Diffusion Schr\"odinger Bridge (DSB) that facilitates its unification with Score-based Generative Models (SGMs), addressing the limitations of DSB in complex data generation and enabling faster convergence and enhanced performance. By employing SGMs as an initial solution for DSB, our approach capitalizes on the strengths of both frameworks, ensuring a more efficient training process and improving the performance of SGM. We also propose a reparameterization technique that, despite theoretical approximations, practically improves the network's fitting capabilities. Our extensive experimental evaluations confirm the effectiveness of the simplified DSB, demonstrating its significant improvements. We believe the contributions of this work pave the way for advanced generative modeling.

1511.04376

Martin Reisslein

Akhilesh Thyagaturu, Anu Mercian, Michael P. McGarry, Martin Reisslein, Wolfgang Kellerer

Software Defined Optical Networks (SDONs): A Comprehensive Survey

null

IEEE Communications Surveys & Tutorials, vol. 18, no. 4, pp. 2738-2786, 4th Qu. 2016

10.1109/COMST.2016.2586999

null

cs.NI

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The emerging Software Defined Networking (SDN) paradigm separates the data plane from the control plane and centralizes network control in an SDN controller. Applications interact with controllers to implement network services, such as network transport with Quality of Service (QoS). SDN facilitates the virtualization of network functions so that multiple virtual networks can operate over a given installed physical network infrastructure. Due to the specific characteristics of optical (photonic) communication components and the high optical transmission capacities, SDN based optical networking poses particular challenges, but holds also great potential. In this article, we comprehensively survey studies that examine the SDN paradigm in optical networks; in brief, we survey the area of Software Defined Optical Networks (SDONs). We mainly organize the SDON studies into studies focused on the infrastructure layer, the control layer, and the application layer. Moreover, we cover SDON studies focused on network virtualization, as well as SDON studies focused on the orchestration of multilayer and multidomain networking. Based on the survey, we identify open challenges for SDONs and outline future directions.

[ { "created": "Fri, 13 Nov 2015 17:31:10 GMT", "version": "v1" }, { "created": "Thu, 26 May 2016 10:19:58 GMT", "version": "v2" }, { "created": "Sun, 17 Jul 2016 07:46:47 GMT", "version": "v3" } ]

2016-11-29

[ [ "Thyagaturu", "Akhilesh", "" ], [ "Mercian", "Anu", "" ], [ "McGarry", "Michael P.", "" ], [ "Reisslein", "Martin", "" ], [ "Kellerer", "Wolfgang", "" ] ]

The emerging Software Defined Networking (SDN) paradigm separates the data plane from the control plane and centralizes network control in an SDN controller. Applications interact with controllers to implement network services, such as network transport with Quality of Service (QoS). SDN facilitates the virtualization of network functions so that multiple virtual networks can operate over a given installed physical network infrastructure. Due to the specific characteristics of optical (photonic) communication components and the high optical transmission capacities, SDN based optical networking poses particular challenges, but holds also great potential. In this article, we comprehensively survey studies that examine the SDN paradigm in optical networks; in brief, we survey the area of Software Defined Optical Networks (SDONs). We mainly organize the SDON studies into studies focused on the infrastructure layer, the control layer, and the application layer. Moreover, we cover SDON studies focused on network virtualization, as well as SDON studies focused on the orchestration of multilayer and multidomain networking. Based on the survey, we identify open challenges for SDONs and outline future directions.

2306.02346

Shuo Ye

Shuo Ye and Yufeng Shi and Ruxin Wang and Yu Wang and Jiamiao Xu and Chuanwu Yang and Xinge You

CDLT: A Dataset with Concept Drift and Long-Tailed Distribution for Fine-Grained Visual Categorization

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Data is the foundation for the development of computer vision, and the establishment of datasets plays an important role in advancing the techniques of fine-grained visual categorization~(FGVC). In the existing FGVC datasets used in computer vision, it is generally assumed that each collected instance has fixed characteristics and the distribution of different categories is relatively balanced. In contrast, the real world scenario reveals the fact that the characteristics of instances tend to vary with time and exhibit a long-tailed distribution. Hence, the collected datasets may mislead the optimization of the fine-grained classifiers, resulting in unpleasant performance in real applications. Starting from the real-world conditions and to promote the practical progress of fine-grained visual categorization, we present a Concept Drift and Long-Tailed Distribution dataset. Specifically, the dataset is collected by gathering 11195 images of 250 instances in different species for 47 consecutive months in their natural contexts. The collection process involves dozens of crowd workers for photographing and domain experts for labelling. Extensive baseline experiments using the state-of-the-art fine-grained classification models demonstrate the issues of concept drift and long-tailed distribution existed in the dataset, which require the attention of future researches.

[ { "created": "Sun, 4 Jun 2023 12:42:45 GMT", "version": "v1" } ]

2023-06-06

[ [ "Ye", "Shuo", "" ], [ "Shi", "Yufeng", "" ], [ "Wang", "Ruxin", "" ], [ "Wang", "Yu", "" ], [ "Xu", "Jiamiao", "" ], [ "Yang", "Chuanwu", "" ], [ "You", "Xinge", "" ] ]

Data is the foundation for the development of computer vision, and the establishment of datasets plays an important role in advancing the techniques of fine-grained visual categorization~(FGVC). In the existing FGVC datasets used in computer vision, it is generally assumed that each collected instance has fixed characteristics and the distribution of different categories is relatively balanced. In contrast, the real world scenario reveals the fact that the characteristics of instances tend to vary with time and exhibit a long-tailed distribution. Hence, the collected datasets may mislead the optimization of the fine-grained classifiers, resulting in unpleasant performance in real applications. Starting from the real-world conditions and to promote the practical progress of fine-grained visual categorization, we present a Concept Drift and Long-Tailed Distribution dataset. Specifically, the dataset is collected by gathering 11195 images of 250 instances in different species for 47 consecutive months in their natural contexts. The collection process involves dozens of crowd workers for photographing and domain experts for labelling. Extensive baseline experiments using the state-of-the-art fine-grained classification models demonstrate the issues of concept drift and long-tailed distribution existed in the dataset, which require the attention of future researches.

2303.06545

Hao Zhou

Hao Zhou, Chongyang Zhang, Yanjun Chen, Chuanping Hu

Towards Diverse Temporal Grounding under Single Positive Labels

The source codes are available at https://github.com/zhouhaocv/DTG-SPL

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Temporal grounding aims to retrieve moments of the described event within an untrimmed video by a language query. Typically, existing methods assume annotations are precise and unique, yet one query may describe multiple moments in many cases. Hence, simply taking it as a one-vs-one mapping task and striving to match single-label annotations will inevitably introduce false negatives during optimization. In this study, we reformulate this task as a one-vs-many optimization problem under the condition of single positive labels. The unlabeled moments are considered unobserved rather than negative, and we explore mining potential positive moments to assist in multiple moment retrieval. In this setting, we propose a novel Diverse Temporal Grounding framework, termed DTG-SPL, which mainly consists of a positive moment estimation (PME) module and a diverse moment regression (DMR) module. PME leverages semantic reconstruction information and an expected positive regularization to uncover potential positive moments in an online fashion. Under the supervision of these pseudo positives, DMR is able to localize diverse moments in parallel that meet different users. The entire framework allows for end-to-end optimization as well as fast inference. Extensive experiments on Charades-STA and ActivityNet Captions show that our method achieves superior performance in terms of both single-label and multi-label metrics.

[ { "created": "Sun, 12 Mar 2023 02:54:18 GMT", "version": "v1" } ]

2023-03-14

[ [ "Zhou", "Hao", "" ], [ "Zhang", "Chongyang", "" ], [ "Chen", "Yanjun", "" ], [ "Hu", "Chuanping", "" ] ]

Temporal grounding aims to retrieve moments of the described event within an untrimmed video by a language query. Typically, existing methods assume annotations are precise and unique, yet one query may describe multiple moments in many cases. Hence, simply taking it as a one-vs-one mapping task and striving to match single-label annotations will inevitably introduce false negatives during optimization. In this study, we reformulate this task as a one-vs-many optimization problem under the condition of single positive labels. The unlabeled moments are considered unobserved rather than negative, and we explore mining potential positive moments to assist in multiple moment retrieval. In this setting, we propose a novel Diverse Temporal Grounding framework, termed DTG-SPL, which mainly consists of a positive moment estimation (PME) module and a diverse moment regression (DMR) module. PME leverages semantic reconstruction information and an expected positive regularization to uncover potential positive moments in an online fashion. Under the supervision of these pseudo positives, DMR is able to localize diverse moments in parallel that meet different users. The entire framework allows for end-to-end optimization as well as fast inference. Extensive experiments on Charades-STA and ActivityNet Captions show that our method achieves superior performance in terms of both single-label and multi-label metrics.

2110.15919

Pranay Bhardwaj

Vinay U. Pai, Pranay Bhardwaj, and S. M. Zafaruddin

Performance Analysis of Dual-Hop THz Wireless Transmission for Backhaul Applications

This paper has been accepted for presentation in 2021 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS), Hyderabad, India

null

cs.IT eess.SP math.IT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

THz transmissions suffer from pointing errors due to antenna misalignment and incur higher path loss from the molecular absorption in addition to the channel fading. In this paper, we employ an amplify-and-forward (AF) dual-hop relaying to mitigate the effect of pointing errors and extend the range of the THz wireless system for backhaul connectivity. We provide statistical analysis on the performance of the considered system by deriving analytical expressions for the outage probability, average bit-error-rate (BER), average signal-to-noise ratio (SNR), and a lower bound on the ergodic capacity over independent and identical (i.i.d) $\alpha$-$\mu$ fading combined with the statistical effect of pointing errors. Using computer simulations, we validate the derived analysis of the relay-assisted system. We also demonstrate the effect of the system parameters on outage probability and average BER with the help of diversity order. We show that data rates up to several \mbox{Gbps} can be achieved using THz transmissions, which is desirable for next-generation wireless systems, especially for backhaul applications.

[ { "created": "Fri, 29 Oct 2021 17:15:38 GMT", "version": "v1" }, { "created": "Sun, 21 Nov 2021 19:24:23 GMT", "version": "v2" } ]

2021-11-23

[ [ "Pai", "Vinay U.", "" ], [ "Bhardwaj", "Pranay", "" ], [ "Zafaruddin", "S. M.", "" ] ]

THz transmissions suffer from pointing errors due to antenna misalignment and incur higher path loss from the molecular absorption in addition to the channel fading. In this paper, we employ an amplify-and-forward (AF) dual-hop relaying to mitigate the effect of pointing errors and extend the range of the THz wireless system for backhaul connectivity. We provide statistical analysis on the performance of the considered system by deriving analytical expressions for the outage probability, average bit-error-rate (BER), average signal-to-noise ratio (SNR), and a lower bound on the ergodic capacity over independent and identical (i.i.d) $\alpha$-$\mu$ fading combined with the statistical effect of pointing errors. Using computer simulations, we validate the derived analysis of the relay-assisted system. We also demonstrate the effect of the system parameters on outage probability and average BER with the help of diversity order. We show that data rates up to several \mbox{Gbps} can be achieved using THz transmissions, which is desirable for next-generation wireless systems, especially for backhaul applications.

2202.02071

Henrique Moniz

Afonso Oliveira, Henrique Moniz, Rodrigo Rodrigues

Alea-BFT: Practical Asynchronous Byzantine Fault Tolerance

null

cs.DC

http://creativecommons.org/licenses/by/4.0/

Traditional Byzantine Fault Tolerance (BFT) state machine replication protocols assume a partial synchrony model, leading to a design where a leader replica drives the protocol and is replaced after a timeout. Recently, we witnessed a surge of asynchronous BFT protocols that use randomization to remove the assumptions of bounds on message delivery times, making them more resilient to adverse network conditions. However, these protocols still fall short of being practical across a broad range of scenarios due to their cubic communication costs, use of expensive primitives, and overall protocol complexity. In this paper, we present Alea-BFT, the first asynchronous BFT protocol to achieve quadratic communication complexity, allowing it to scale to large networks. Alea-BFT brings the key design insight from classical protocols of concentrating part of the work on a single designated replica, and incorporates this principle in a two stage pipelined design, with an efficient broadcast led by the designated replica followed by an inexpensive binary agreement. We evaluated our prototype implementation across 10 sites in 4 continents, and our results show significant scalability gains from the proposed design.

[ { "created": "Fri, 4 Feb 2022 10:53:37 GMT", "version": "v1" } ]

2022-02-07

[ [ "Oliveira", "Afonso", "" ], [ "Moniz", "Henrique", "" ], [ "Rodrigues", "Rodrigo", "" ] ]

Traditional Byzantine Fault Tolerance (BFT) state machine replication protocols assume a partial synchrony model, leading to a design where a leader replica drives the protocol and is replaced after a timeout. Recently, we witnessed a surge of asynchronous BFT protocols that use randomization to remove the assumptions of bounds on message delivery times, making them more resilient to adverse network conditions. However, these protocols still fall short of being practical across a broad range of scenarios due to their cubic communication costs, use of expensive primitives, and overall protocol complexity. In this paper, we present Alea-BFT, the first asynchronous BFT protocol to achieve quadratic communication complexity, allowing it to scale to large networks. Alea-BFT brings the key design insight from classical protocols of concentrating part of the work on a single designated replica, and incorporates this principle in a two stage pipelined design, with an efficient broadcast led by the designated replica followed by an inexpensive binary agreement. We evaluated our prototype implementation across 10 sites in 4 continents, and our results show significant scalability gains from the proposed design.

1405.2066

Jehad Al Dallal Prof.

Jehad Al Dallal

How and When to Flatten Java Classes?

null

cs.SE

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Improving modularity and reusability are two key objectives in object-oriented programming. These objectives are achieved by applying several key concepts, such as data encapsulation and inheritance. A class in an object-oriented system is the basic unit of design. Assessing the quality of an object-oriented class may require flattening the class and representing it as it really is, including all accessible inherited class members. Thus, class flattening helps in exploring the impact of inheritance on improving code quality. This paper explains how to flatten Java classes and discusses the relationship between class flattening and some applications of interest to software practitioners, such as refactoring and indicating external quality attributes.

[ { "created": "Thu, 8 May 2014 19:48:33 GMT", "version": "v1" } ]

2014-05-09

[ [ "Dallal", "Jehad Al", "" ] ]

Improving modularity and reusability are two key objectives in object-oriented programming. These objectives are achieved by applying several key concepts, such as data encapsulation and inheritance. A class in an object-oriented system is the basic unit of design. Assessing the quality of an object-oriented class may require flattening the class and representing it as it really is, including all accessible inherited class members. Thus, class flattening helps in exploring the impact of inheritance on improving code quality. This paper explains how to flatten Java classes and discusses the relationship between class flattening and some applications of interest to software practitioners, such as refactoring and indicating external quality attributes.

1406.3583

Aaron D. Jaggard

Aaron D. Jaggard, Aaron Johnson, Paul Syverson, and Joan Feigenbaum

Representing Network Trust and Using It to Improve Anonymous Communication

24 pages; talk to be presented at HotPETs 2014

null

cs.CR

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Motivated by the effectiveness of correlation attacks against Tor, the censorship arms race, and observations of malicious relays in Tor, we propose that Tor users capture their trust in network elements using probability distributions over the sets of elements observed by network adversaries. We present a modular system that allows users to efficiently and conveniently create such distributions and use them to improve their security. The major components of this system are (i) an ontology of network-element types that represents the main threats to and vulnerabilities of anonymous communication over Tor, (ii) a formal language that allows users to naturally express trust beliefs about network elements, and (iii) a conversion procedure that takes the ontology, public information about the network, and user beliefs written in the trust language and produce a Bayesian Belief Network that represents the probability distribution in a way that is concise and easily sampleable. We also present preliminary experimental results that show the distribution produced by our system can improve security when employed by users; further improvement is seen when the system is employed by both users and services.

[ { "created": "Fri, 13 Jun 2014 16:23:12 GMT", "version": "v1" } ]

2014-06-16

[ [ "Jaggard", "Aaron D.", "" ], [ "Johnson", "Aaron", "" ], [ "Syverson", "Paul", "" ], [ "Feigenbaum", "Joan", "" ] ]

Motivated by the effectiveness of correlation attacks against Tor, the censorship arms race, and observations of malicious relays in Tor, we propose that Tor users capture their trust in network elements using probability distributions over the sets of elements observed by network adversaries. We present a modular system that allows users to efficiently and conveniently create such distributions and use them to improve their security. The major components of this system are (i) an ontology of network-element types that represents the main threats to and vulnerabilities of anonymous communication over Tor, (ii) a formal language that allows users to naturally express trust beliefs about network elements, and (iii) a conversion procedure that takes the ontology, public information about the network, and user beliefs written in the trust language and produce a Bayesian Belief Network that represents the probability distribution in a way that is concise and easily sampleable. We also present preliminary experimental results that show the distribution produced by our system can improve security when employed by users; further improvement is seen when the system is employed by both users and services.

1702.06235

Will Radford

Andrew Chisholm, Will Radford, Ben Hachey

Learning to generate one-sentence biographies from Wikidata

null

cs.CL

http://creativecommons.org/licenses/by/4.0/

We investigate the generation of one-sentence Wikipedia biographies from facts derived from Wikidata slot-value pairs. We train a recurrent neural network sequence-to-sequence model with attention to select facts and generate textual summaries. Our model incorporates a novel secondary objective that helps ensure it generates sentences that contain the input facts. The model achieves a BLEU score of 41, improving significantly upon the vanilla sequence-to-sequence model and scoring roughly twice that of a simple template baseline. Human preference evaluation suggests the model is nearly as good as the Wikipedia reference. Manual analysis explores content selection, suggesting the model can trade the ability to infer knowledge against the risk of hallucinating incorrect information.

[ { "created": "Tue, 21 Feb 2017 01:30:59 GMT", "version": "v1" } ]

2017-02-22

[ [ "Chisholm", "Andrew", "" ], [ "Radford", "Will", "" ], [ "Hachey", "Ben", "" ] ]

We investigate the generation of one-sentence Wikipedia biographies from facts derived from Wikidata slot-value pairs. We train a recurrent neural network sequence-to-sequence model with attention to select facts and generate textual summaries. Our model incorporates a novel secondary objective that helps ensure it generates sentences that contain the input facts. The model achieves a BLEU score of 41, improving significantly upon the vanilla sequence-to-sequence model and scoring roughly twice that of a simple template baseline. Human preference evaluation suggests the model is nearly as good as the Wikipedia reference. Manual analysis explores content selection, suggesting the model can trade the ability to infer knowledge against the risk of hallucinating incorrect information.

2402.11173

Andrew Lowy

Andrew Lowy, Jonathan Ullman, Stephen J. Wright

How to Make the Gradients Small Privately: Improved Rates for Differentially Private Non-Convex Optimization

null

cs.LG cs.CR math.OC

http://creativecommons.org/licenses/by/4.0/

We provide a simple and flexible framework for designing differentially private algorithms to find approximate stationary points of non-convex loss functions. Our framework is based on using a private approximate risk minimizer to "warm start" another private algorithm for finding stationary points. We use this framework to obtain improved, and sometimes optimal, rates for several classes of non-convex loss functions. First, we obtain improved rates for finding stationary points of smooth non-convex empirical loss functions. Second, we specialize to quasar-convex functions, which generalize star-convex functions and arise in learning dynamical systems and training some neural nets. We achieve the optimal rate for this class. Third, we give an optimal algorithm for finding stationary points of functions satisfying the Kurdyka-Lojasiewicz (KL) condition. For example, over-parameterized neural networks often satisfy this condition. Fourth, we provide new state-of-the-art rates for stationary points of non-convex population loss functions. Fifth, we obtain improved rates for non-convex generalized linear models. A modification of our algorithm achieves nearly the same rates for second-order stationary points of functions with Lipschitz Hessian, improving over the previous state-of-the-art for each of the above problems.

[ { "created": "Sat, 17 Feb 2024 02:42:56 GMT", "version": "v1" } ]

2024-02-20

[ [ "Lowy", "Andrew", "" ], [ "Ullman", "Jonathan", "" ], [ "Wright", "Stephen J.", "" ] ]

We provide a simple and flexible framework for designing differentially private algorithms to find approximate stationary points of non-convex loss functions. Our framework is based on using a private approximate risk minimizer to "warm start" another private algorithm for finding stationary points. We use this framework to obtain improved, and sometimes optimal, rates for several classes of non-convex loss functions. First, we obtain improved rates for finding stationary points of smooth non-convex empirical loss functions. Second, we specialize to quasar-convex functions, which generalize star-convex functions and arise in learning dynamical systems and training some neural nets. We achieve the optimal rate for this class. Third, we give an optimal algorithm for finding stationary points of functions satisfying the Kurdyka-Lojasiewicz (KL) condition. For example, over-parameterized neural networks often satisfy this condition. Fourth, we provide new state-of-the-art rates for stationary points of non-convex population loss functions. Fifth, we obtain improved rates for non-convex generalized linear models. A modification of our algorithm achieves nearly the same rates for second-order stationary points of functions with Lipschitz Hessian, improving over the previous state-of-the-art for each of the above problems.

2401.11946

Jiajun Liu

Jiajun Liu, Lina Tan, Zhili Zhou, Yi Li, Peng Chen

A Dynamic YOLO-Based Sequence-Matching Model for Efficient Coverless Image Steganography

null

cs.CR

http://creativecommons.org/licenses/by/4.0/

Many existing coverless steganography methods establish a mapping relationship between cover images and hidden data. There exists an issue that the number of images stored in the database grows exponentially as the steganographic capacity rises. The need for a high steganographic capacity makes it challenging to build an image database. To improve the image library utilization and anti-attack capability of the steganography system, we present an efficient coverless scheme based on dynamically matched substrings. YOLO is employed for selecting optimal objects, and a mapping dictionary is established between these objects and scrambling factors. With the aid of this dictionary, each image is effectively assigned to a specific scrambling factor, which is used to scramble the receiver's sequence key. To achieve sufficient steganography capability based on a limited image library, all substrings of the scrambled sequences hold the potential to hide data. After completing the secret information matching, the ideal number of stego images will be obtained from the database. According to experimental results, this technology outperforms most previous works on data load, transmission security, and hiding capacity. Under typical geometric attacks, it can recover 79.85\% of secret information on average. Furthermore, only approximately 200 random images are needed to meet a capacity of 19 bits per image.

[ { "created": "Mon, 22 Jan 2024 13:35:27 GMT", "version": "v1" } ]

2024-01-23

[ [ "Liu", "Jiajun", "" ], [ "Tan", "Lina", "" ], [ "Zhou", "Zhili", "" ], [ "Li", "Yi", "" ], [ "Chen", "Peng", "" ] ]

Many existing coverless steganography methods establish a mapping relationship between cover images and hidden data. There exists an issue that the number of images stored in the database grows exponentially as the steganographic capacity rises. The need for a high steganographic capacity makes it challenging to build an image database. To improve the image library utilization and anti-attack capability of the steganography system, we present an efficient coverless scheme based on dynamically matched substrings. YOLO is employed for selecting optimal objects, and a mapping dictionary is established between these objects and scrambling factors. With the aid of this dictionary, each image is effectively assigned to a specific scrambling factor, which is used to scramble the receiver's sequence key. To achieve sufficient steganography capability based on a limited image library, all substrings of the scrambled sequences hold the potential to hide data. After completing the secret information matching, the ideal number of stego images will be obtained from the database. According to experimental results, this technology outperforms most previous works on data load, transmission security, and hiding capacity. Under typical geometric attacks, it can recover 79.85\% of secret information on average. Furthermore, only approximately 200 random images are needed to meet a capacity of 19 bits per image.

2203.10837

Marcos Faundez-Zanuy

K. L\'opez-de-Ipi\~na, Marcos Faundez-Zanuy, Jordi Sol\'e-Casals, Fernando Zelarin, Pilar Calvo

Multi-class versus One-class classifier in spontaneous speech analysis oriented to Alzheimer Disease diagnosis

10 pages, published in International Conference on NONLINEAR SPEECH PROCESSING, NOLISP 2015 jointly organized with the 25th Italian Workshop on Neural Networks, WIRN 2015, held at May 2015, Vietri sul Mare, Salerno, Italy

Recent Advances in Nonlinear Speech Processing. Smart Innovation, Systems and Technologies, vol 48. Springer, Cham 2015

10.1007/978-3-319-28109-4_7

null

cs.SD cs.LG eess.AS

http://creativecommons.org/licenses/by-nc-nd/4.0/

Most of medical developments require the ability to identify samples that are anomalous with respect to a target group or control group, in the sense they could belong to a new, previously unseen class or are not class data. In this case when there are not enough data to train two-class One-class classification appear like an available solution. On the other hand non-linear approaches could give very useful information. The aim of our project is to contribute to earlier diagnosis of AD and better estimates of its severity by using automatic analysis performed through new biomarkers extracted from speech signal. The methods selected in this case are speech biomarkers oriented to Spontaneous Speech and Emotional Response Analysis. In this approach One-class classifiers and two-class classifiers are analyzed. The use of information about outlier and Fractal Dimension features improves the system performance.

[ { "created": "Mon, 21 Mar 2022 09:57:20 GMT", "version": "v1" } ]

2022-03-22

[ [ "López-de-Ipiña", "K.", "" ], [ "Faundez-Zanuy", "Marcos", "" ], [ "Solé-Casals", "Jordi", "" ], [ "Zelarin", "Fernando", "" ], [ "Calvo", "Pilar", "" ] ]

Most of medical developments require the ability to identify samples that are anomalous with respect to a target group or control group, in the sense they could belong to a new, previously unseen class or are not class data. In this case when there are not enough data to train two-class One-class classification appear like an available solution. On the other hand non-linear approaches could give very useful information. The aim of our project is to contribute to earlier diagnosis of AD and better estimates of its severity by using automatic analysis performed through new biomarkers extracted from speech signal. The methods selected in this case are speech biomarkers oriented to Spontaneous Speech and Emotional Response Analysis. In this approach One-class classifiers and two-class classifiers are analyzed. The use of information about outlier and Fractal Dimension features improves the system performance.

2408.06047

Xuanpu Zhang

Xuanpu Zhang and Dan Song and Pengxin Zhan and Qingguo Chen and Zhao Xu and Weihua Luo and Kaifu Zhang and Anan Liu

BooW-VTON: Boosting In-the-Wild Virtual Try-On via Mask-Free Pseudo Data Training

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Image-based virtual try-on is an increasingly popular and important task to generate realistic try-on images of specific person. Existing methods always employ an accurate mask to remove the original garment in the source image, thus achieving realistic synthesized images in simple and conventional try-on scenarios based on powerful diffusion model. Therefore, acquiring suitable mask is vital to the try-on performance of these methods. However, obtaining precise inpainting masks, especially for complex wild try-on data containing diverse foreground occlusions and person poses, is not easy as Figure 1-Top shows. This difficulty often results in poor performance in more practical and challenging real-life scenarios, such as the selfie scene shown in Figure 1-Bottom. To this end, we propose a novel training paradigm combined with an efficient data augmentation method to acquire large-scale unpaired training data from wild scenarios, thereby significantly facilitating the try-on performance of our model without the need for additional inpainting masks. Besides, a try-on localization loss is designed to localize a more accurate try-on area to obtain more reasonable try-on results. It is noted that our method only needs the reference cloth image, source pose image and source person image as input, which is more cost-effective and user-friendly compared to existing methods. Extensive qualitative and quantitative experiments have demonstrated superior performance in wild scenarios with such a low-demand input.

[ { "created": "Mon, 12 Aug 2024 10:39:59 GMT", "version": "v1" } ]

2024-08-13

[ [ "Zhang", "Xuanpu", "" ], [ "Song", "Dan", "" ], [ "Zhan", "Pengxin", "" ], [ "Chen", "Qingguo", "" ], [ "Xu", "Zhao", "" ], [ "Luo", "Weihua", "" ], [ "Zhang", "Kaifu", "" ], [ "Liu", "Anan", "" ] ]

Image-based virtual try-on is an increasingly popular and important task to generate realistic try-on images of specific person. Existing methods always employ an accurate mask to remove the original garment in the source image, thus achieving realistic synthesized images in simple and conventional try-on scenarios based on powerful diffusion model. Therefore, acquiring suitable mask is vital to the try-on performance of these methods. However, obtaining precise inpainting masks, especially for complex wild try-on data containing diverse foreground occlusions and person poses, is not easy as Figure 1-Top shows. This difficulty often results in poor performance in more practical and challenging real-life scenarios, such as the selfie scene shown in Figure 1-Bottom. To this end, we propose a novel training paradigm combined with an efficient data augmentation method to acquire large-scale unpaired training data from wild scenarios, thereby significantly facilitating the try-on performance of our model without the need for additional inpainting masks. Besides, a try-on localization loss is designed to localize a more accurate try-on area to obtain more reasonable try-on results. It is noted that our method only needs the reference cloth image, source pose image and source person image as input, which is more cost-effective and user-friendly compared to existing methods. Extensive qualitative and quantitative experiments have demonstrated superior performance in wild scenarios with such a low-demand input.

1604.01186

EPTCS

Denis Firsov, Tarmo Uustalu, Niccol\`o Veltri

Variations on Noetherianness

In Proceedings MSFP 2016, arXiv:1604.00384

EPTCS 207, 2016, pp. 76-88

10.4204/EPTCS.207.4

null

cs.LO cs.PL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In constructive mathematics, several nonequivalent notions of finiteness exist. In this paper, we continue the study of Noetherian sets in the dependently typed setting of the Agda programming language. We want to say that a set is Noetherian, if, when we are shown elements from it one after another, we will sooner or later have seen some element twice. This idea can be made precise in a number of ways. We explore the properties and connections of some of the possible encodings. In particular, we show that certain implementations imply decidable equality while others do not, and we construct counterexamples in the latter case. Additionally, we explore the relation between Noetherianness and other notions of finiteness.

[ { "created": "Tue, 5 Apr 2016 09:04:13 GMT", "version": "v1" } ]

2016-04-06

[ [ "Firsov", "Denis", "" ], [ "Uustalu", "Tarmo", "" ], [ "Veltri", "Niccolò", "" ] ]

In constructive mathematics, several nonequivalent notions of finiteness exist. In this paper, we continue the study of Noetherian sets in the dependently typed setting of the Agda programming language. We want to say that a set is Noetherian, if, when we are shown elements from it one after another, we will sooner or later have seen some element twice. This idea can be made precise in a number of ways. We explore the properties and connections of some of the possible encodings. In particular, we show that certain implementations imply decidable equality while others do not, and we construct counterexamples in the latter case. Additionally, we explore the relation between Noetherianness and other notions of finiteness.

2109.06612

Johanna Schmidt

Raphael Sahann, Torsten M\"oller, Johanna Schmidt

Histogram binning revisited with a focus on human perception

Accepted as short paper at VIS 2021. Supplemental material can be found at https://github.com/johanna-schmidt/histogram-binning-revisited

Proceedings of VIS short papers 2021

null

cs.HC

http://creativecommons.org/licenses/by-nc-sa/4.0/

This paper presents a quantitative user study to evaluate how well users can visually perceive the underlying data distribution from a histogram representation. We used different sample and bin sizes and four different distributions (uniform, normal, bimodal, and gamma). The study results confirm that, in general, more bins correlate with fewer errors by the viewers. However, upon a certain number of bins, the error rate cannot be improved by adding more bins. By comparing our study results with the outcomes of existing mathematical models for histogram binning (e.g., Sturges' formula, Scott's normal reference rule, the Rice Rule, or Freedman-Diaconis' choice), we can see that most of them overestimate the number of bins necessary to make the distribution visible to a human viewer.

[ { "created": "Tue, 14 Sep 2021 12:08:27 GMT", "version": "v1" } ]

2021-09-15

[ [ "Sahann", "Raphael", "" ], [ "Möller", "Torsten", "" ], [ "Schmidt", "Johanna", "" ] ]

This paper presents a quantitative user study to evaluate how well users can visually perceive the underlying data distribution from a histogram representation. We used different sample and bin sizes and four different distributions (uniform, normal, bimodal, and gamma). The study results confirm that, in general, more bins correlate with fewer errors by the viewers. However, upon a certain number of bins, the error rate cannot be improved by adding more bins. By comparing our study results with the outcomes of existing mathematical models for histogram binning (e.g., Sturges' formula, Scott's normal reference rule, the Rice Rule, or Freedman-Diaconis' choice), we can see that most of them overestimate the number of bins necessary to make the distribution visible to a human viewer.

1802.09657

Dipankar Maity

Dipankar Maity, John S. Baras

Event-Triggered Controller Synthesis for Dynamical Systems with Temporal Logic Constraints

null

cs.RO math.DS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this work, we propose an event-triggered con- trol framework for dynamical systems with temporal logical constraints. Event-triggered control methodologies have proven to be very efficient in reducing sensing, communication and computation costs. When a continuous feedback control is re- placed with an event-triggered strategy, the corresponding state trajectories also differ. In a system with logical constraints, such small deviation in the trajectory might lead to unsatisfiability of the logical constraints. In this work, we develop an approach where we ensure that the event-triggered state trajectory is confined within an tube of the ideal trajectory associated with the continuous state feedback. At the same time, we will ensure satisfiability of the logical constraints as well. Furthermore, we show that the proposed method works for delayed systems as long as the delay is bounded by a certain quantity.

[ { "created": "Tue, 27 Feb 2018 00:17:41 GMT", "version": "v1" } ]

2018-02-28

[ [ "Maity", "Dipankar", "" ], [ "Baras", "John S.", "" ] ]

In this work, we propose an event-triggered con- trol framework for dynamical systems with temporal logical constraints. Event-triggered control methodologies have proven to be very efficient in reducing sensing, communication and computation costs. When a continuous feedback control is re- placed with an event-triggered strategy, the corresponding state trajectories also differ. In a system with logical constraints, such small deviation in the trajectory might lead to unsatisfiability of the logical constraints. In this work, we develop an approach where we ensure that the event-triggered state trajectory is confined within an tube of the ideal trajectory associated with the continuous state feedback. At the same time, we will ensure satisfiability of the logical constraints as well. Furthermore, we show that the proposed method works for delayed systems as long as the delay is bounded by a certain quantity.

1911.03059

Chowdhury Rahman

Afra Anika, Md. Hasibur Rahman, Salekul Islam, Abu Shafin Mohammad Mahdee Jameel and Chowdhury Rafeed Rahman

A Comprehensive Comparison of Machine Learning Based Methods Used in Bengali Question Classification

null

cs.CL

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

QA classification system maps questions asked by humans to an appropriate answer category. A sound question classification (QC) system model is the pre-requisite of a sound QA system. This work demonstrates phases of assembling a QA type classification model. We present a comprehensive comparison (performance and computational complexity) among some machine learning based approaches used in QC for Bengali language.

[ { "created": "Fri, 8 Nov 2019 05:30:33 GMT", "version": "v1" }, { "created": "Tue, 19 Nov 2019 16:37:41 GMT", "version": "v2" } ]

2019-11-20

[ [ "Anika", "Afra", "" ], [ "Rahman", "Md. Hasibur", "" ], [ "Islam", "Salekul", "" ], [ "Jameel", "Abu Shafin Mohammad Mahdee", "" ], [ "Rahman", "Chowdhury Rafeed", "" ] ]

QA classification system maps questions asked by humans to an appropriate answer category. A sound question classification (QC) system model is the pre-requisite of a sound QA system. This work demonstrates phases of assembling a QA type classification model. We present a comprehensive comparison (performance and computational complexity) among some machine learning based approaches used in QC for Bengali language.

2212.09448

Senem Tanberk PhD

Senem Tanberk, Mustafa Can

Smart Journey in Istanbul: A Mobile Application in Smart Cities for Traffic Estimation by Harnessing Time Series

null

10.1109/ASYU58738.2023.10296669

null

cs.AI cs.IR

http://creativecommons.org/licenses/by/4.0/

In recent decades, mobile applications (apps) have gained enormous popularity. Smart services for smart cities increasingly gain attention. The main goal of the proposed research is to present a new AI-powered mobile application on Istanbul's traffic congestion forecast by using traffic density data. It addresses the research question by using time series approaches (LSTM, Transformer, and XGBoost) based on past data over the traffic load dataset combined with meteorological conditions. Analysis of simulation results on predicted models will be discussed according to performance indicators such as MAPE, MAE, and RMSE. And then, it was observed that the Transformer model made the most accurate traffic prediction. The developed traffic forecasting prototype is expected to be a starting point on future products for a mobile application suitable for citizens' daily use.

[ { "created": "Tue, 13 Dec 2022 12:10:52 GMT", "version": "v1" } ]

2023-11-09

[ [ "Tanberk", "Senem", "" ], [ "Can", "Mustafa", "" ] ]

In recent decades, mobile applications (apps) have gained enormous popularity. Smart services for smart cities increasingly gain attention. The main goal of the proposed research is to present a new AI-powered mobile application on Istanbul's traffic congestion forecast by using traffic density data. It addresses the research question by using time series approaches (LSTM, Transformer, and XGBoost) based on past data over the traffic load dataset combined with meteorological conditions. Analysis of simulation results on predicted models will be discussed according to performance indicators such as MAPE, MAE, and RMSE. And then, it was observed that the Transformer model made the most accurate traffic prediction. The developed traffic forecasting prototype is expected to be a starting point on future products for a mobile application suitable for citizens' daily use.

2401.11150

Junxiao Shen Mr

Junxiao Shen, Xuhai Xu, Ran Tan, Amy Karlson, Evan Strasnick

Simultaneous Gesture Classification and Localization with an Automatic Gesture Annotation Model

null

cs.CV

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Training a real-time gesture recognition model heavily relies on annotated data. However, manual data annotation is costly and demands substantial human effort. In order to address this challenge, we propose a novel annotation model that can automatically annotate gesture classes and identify their temporal ranges. Our ablation study demonstrates that our annotation model design surpasses the baseline in terms of both gesture classification accuracy (3-4\% improvement) and localization accuracy (71-75\% improvement). We believe that this annotation model has immense potential to improve the training of downstream gesture recognition models using unlabeled datasets.

[ { "created": "Sat, 20 Jan 2024 07:11:03 GMT", "version": "v1" } ]

2024-01-23

[ [ "Shen", "Junxiao", "" ], [ "Xu", "Xuhai", "" ], [ "Tan", "Ran", "" ], [ "Karlson", "Amy", "" ], [ "Strasnick", "Evan", "" ] ]

Training a real-time gesture recognition model heavily relies on annotated data. However, manual data annotation is costly and demands substantial human effort. In order to address this challenge, we propose a novel annotation model that can automatically annotate gesture classes and identify their temporal ranges. Our ablation study demonstrates that our annotation model design surpasses the baseline in terms of both gesture classification accuracy (3-4\% improvement) and localization accuracy (71-75\% improvement). We believe that this annotation model has immense potential to improve the training of downstream gesture recognition models using unlabeled datasets.

1005.0907

Rdv Ijcsis

Yasser M. Alginaih, Abdul Ahad Siddiqi

Multistage Hybrid Arabic/Indian Numeral OCR System

IEEE Publication format, International Journal of Computer Science and Information Security, IJCSIS, Vol. 8 No. 1, April 2010, USA. ISSN 1947 5500, http://sites.google.com/site/ijcsis/

null

cs.CV

http://creativecommons.org/licenses/by-nc-sa/3.0/

The use of OCR in postal services is not yet universal and there are still many countries that process mail sorting manually. Automated Arabic/Indian numeral Optical Character Recognition (OCR) systems for Postal services are being used in some countries, but still there are errors during the mail sorting process, thus causing a reduction in efficiency. The need to investigate fast and efficient recognition algorithms/systems is important so as to correctly read the postal codes from mail addresses and to eliminate any errors during the mail sorting stage. The objective of this study is to recognize printed numerical postal codes from mail addresses. The proposed system is a multistage hybrid system which consists of three different feature extraction methods, i.e., binary, zoning, and fuzzy features, and three different classifiers, i.e., Hamming Nets, Euclidean Distance, and Fuzzy Neural Network Classifiers. The proposed system, systematically compares the performance of each of these methods, and ensures that the numerals are recognized correctly. Comprehensive results provide a very high recognition rate, outperforming the other known developed methods in literature.

[ { "created": "Thu, 6 May 2010 07:25:23 GMT", "version": "v1" } ]

2010-05-07

[ [ "Alginaih", "Yasser M.", "" ], [ "Siddiqi", "Abdul Ahad", "" ] ]

The use of OCR in postal services is not yet universal and there are still many countries that process mail sorting manually. Automated Arabic/Indian numeral Optical Character Recognition (OCR) systems for Postal services are being used in some countries, but still there are errors during the mail sorting process, thus causing a reduction in efficiency. The need to investigate fast and efficient recognition algorithms/systems is important so as to correctly read the postal codes from mail addresses and to eliminate any errors during the mail sorting stage. The objective of this study is to recognize printed numerical postal codes from mail addresses. The proposed system is a multistage hybrid system which consists of three different feature extraction methods, i.e., binary, zoning, and fuzzy features, and three different classifiers, i.e., Hamming Nets, Euclidean Distance, and Fuzzy Neural Network Classifiers. The proposed system, systematically compares the performance of each of these methods, and ensures that the numerals are recognized correctly. Comprehensive results provide a very high recognition rate, outperforming the other known developed methods in literature.

2405.13937

Xingtong Yu

Xingtong Yu, Zhenghao Liu, Yuan Fang, Xinming Zhang

DyGPrompt: Learning Feature and Time Prompts on Dynamic Graphs

Under review

null

cs.LG

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Dynamic graphs are pervasive in the real world, modeling dynamic relations between objects across various fields. For dynamic graph modeling, dynamic graph neural networks (DGNNs) have emerged as a mainstream technique, which are generally pre-trained on the link prediction task, leaving a significant gap from the objectives of downstream tasks such as node classification. To bridge the gap, prompt-based learning has gained traction on graphs. However, existing efforts focus on static graphs, neglecting the evolution of dynamic graphs. In this paper, we propose DyGPrompt, a novel pre-training and prompting framework for dynamic graph modeling. First, we design dual prompts to address the gap in both task objectives and dynamic variations across pre-training and downstream tasks. Second, we recognize that node and time features mutually characterize each other, and propose dual condition-nets to model the evolving node-time patterns in downstream tasks. Finally, we thoroughly evaluate and analyze DyGPrompt through extensive experiments on three public datasets.

[ { "created": "Wed, 22 May 2024 19:10:24 GMT", "version": "v1" }, { "created": "Sun, 26 May 2024 01:46:11 GMT", "version": "v2" }, { "created": "Tue, 28 May 2024 10:07:29 GMT", "version": "v3" }, { "created": "Tue, 2 Jul 2024 05:14:10 GMT", "version": "v4" }, { "created": "Wed, 3 Jul 2024 02:06:07 GMT", "version": "v5" } ]

2024-07-04

[ [ "Yu", "Xingtong", "" ], [ "Liu", "Zhenghao", "" ], [ "Fang", "Yuan", "" ], [ "Zhang", "Xinming", "" ] ]

Dynamic graphs are pervasive in the real world, modeling dynamic relations between objects across various fields. For dynamic graph modeling, dynamic graph neural networks (DGNNs) have emerged as a mainstream technique, which are generally pre-trained on the link prediction task, leaving a significant gap from the objectives of downstream tasks such as node classification. To bridge the gap, prompt-based learning has gained traction on graphs. However, existing efforts focus on static graphs, neglecting the evolution of dynamic graphs. In this paper, we propose DyGPrompt, a novel pre-training and prompting framework for dynamic graph modeling. First, we design dual prompts to address the gap in both task objectives and dynamic variations across pre-training and downstream tasks. Second, we recognize that node and time features mutually characterize each other, and propose dual condition-nets to model the evolving node-time patterns in downstream tasks. Finally, we thoroughly evaluate and analyze DyGPrompt through extensive experiments on three public datasets.

2208.02884

Nicolaas Kaashoek

Nicolaas Kaashoek and Robert Morris

CheckSync: Using Runtime-Integrated Checkpoints to Achieve High Availability}

14 pages, 6 figures

null

cs.DC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

CheckSync provides applications with high availability via runtime-integrated checkpointing. This allows CheckSync to take checkpoints of a process running in a memory-managed language (Go, for now), which can be resumed on another machine after a failure. CheckSync uses the runtime to checkpoint only the process' live memory, doing without requiring significant changes to applications. CheckSync maintains the ease of use provided by virtual machines for the applications it supports without requiring that an entire virtual machine image be snapshotted. Because CheckSync captures only the memory used by an application, it produces checkpoints that are smaller (by an order of magnitude) than virtual machine snapshots if the memory footprint of the application is relatively small compared to the state of the rest of the operating system. Additionally, when running go-cache, a popular in-memory key/value store, CheckSync reduces throughput by only 12% compared to the 78% throughput loss when using go-cache's snapshot functionality, the 45% loss when using CRIU, and the 68% loss when using virtual machine live migration.

[ { "created": "Thu, 4 Aug 2022 20:53:50 GMT", "version": "v1" } ]

2022-08-08

[ [ "Kaashoek", "Nicolaas", "" ], [ "Morris", "Robert", "" ] ]

CheckSync provides applications with high availability via runtime-integrated checkpointing. This allows CheckSync to take checkpoints of a process running in a memory-managed language (Go, for now), which can be resumed on another machine after a failure. CheckSync uses the runtime to checkpoint only the process' live memory, doing without requiring significant changes to applications. CheckSync maintains the ease of use provided by virtual machines for the applications it supports without requiring that an entire virtual machine image be snapshotted. Because CheckSync captures only the memory used by an application, it produces checkpoints that are smaller (by an order of magnitude) than virtual machine snapshots if the memory footprint of the application is relatively small compared to the state of the rest of the operating system. Additionally, when running go-cache, a popular in-memory key/value store, CheckSync reduces throughput by only 12% compared to the 78% throughput loss when using go-cache's snapshot functionality, the 45% loss when using CRIU, and the 68% loss when using virtual machine live migration.

2303.13299

Avi Schwarzschild

Avi Schwarzschild, Max Cembalest, Karthik Rao, Keegan Hines, John Dickerson

Reckoning with the Disagreement Problem: Explanation Consensus as a Training Objective

null

cs.LG cs.AI

http://creativecommons.org/licenses/by/4.0/

As neural networks increasingly make critical decisions in high-stakes settings, monitoring and explaining their behavior in an understandable and trustworthy manner is a necessity. One commonly used type of explainer is post hoc feature attribution, a family of methods for giving each feature in an input a score corresponding to its influence on a model's output. A major limitation of this family of explainers in practice is that they can disagree on which features are more important than others. Our contribution in this paper is a method of training models with this disagreement problem in mind. We do this by introducing a Post hoc Explainer Agreement Regularization (PEAR) loss term alongside the standard term corresponding to accuracy, an additional term that measures the difference in feature attribution between a pair of explainers. We observe on three datasets that we can train a model with this loss term to improve explanation consensus on unseen data, and see improved consensus between explainers other than those used in the loss term. We examine the trade-off between improved consensus and model performance. And finally, we study the influence our method has on feature attribution explanations.

[ { "created": "Thu, 23 Mar 2023 14:35:37 GMT", "version": "v1" } ]

2023-03-24

[ [ "Schwarzschild", "Avi", "" ], [ "Cembalest", "Max", "" ], [ "Rao", "Karthik", "" ], [ "Hines", "Keegan", "" ], [ "Dickerson", "John", "" ] ]

As neural networks increasingly make critical decisions in high-stakes settings, monitoring and explaining their behavior in an understandable and trustworthy manner is a necessity. One commonly used type of explainer is post hoc feature attribution, a family of methods for giving each feature in an input a score corresponding to its influence on a model's output. A major limitation of this family of explainers in practice is that they can disagree on which features are more important than others. Our contribution in this paper is a method of training models with this disagreement problem in mind. We do this by introducing a Post hoc Explainer Agreement Regularization (PEAR) loss term alongside the standard term corresponding to accuracy, an additional term that measures the difference in feature attribution between a pair of explainers. We observe on three datasets that we can train a model with this loss term to improve explanation consensus on unseen data, and see improved consensus between explainers other than those used in the loss term. We examine the trade-off between improved consensus and model performance. And finally, we study the influence our method has on feature attribution explanations.

2311.11596

Yining Miao

Yining Miao, Nanlin Shi, Changxing Huang, Yonghao Song, Xiaogang Chen, Yijun Wang, Xiaorong Gao

High-performance cVEP-BCI under minimal calibration

35 pages, 5 figures

null

cs.HC cs.IT eess.SP math.IT q-bio.NC

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

The ultimate goal of brain-computer interfaces (BCIs) based on visual modulation paradigms is to achieve high-speed performance without the burden of extensive calibration. Code-modulated visual evoked potential-based BCIs (cVEP-BCIs) modulated by broadband white noise (WN) offer various advantages, including increased communication speed, expanded encoding target capabilities, and enhanced coding flexibility. However, the complexity of the spatial-temporal patterns under broadband stimuli necessitates extensive calibration for effective target identification in cVEP-BCIs. Consequently, the information transfer rate (ITR) of cVEP-BCI under limited calibration usually stays around 100 bits per minute (bpm), significantly lagging behind state-of-the-art steady-state visual evoked potential-based BCIs (SSVEP-BCIs), which achieve rates above 200 bpm. To enhance the performance of cVEP-BCIs with minimal calibration, we devised an efficient calibration stage involving a brief single-target flickering, lasting less than a minute, to extract generalizable spatial-temporal patterns. Leveraging the calibration data, we developed two complementary methods to construct cVEP temporal patterns: the linear modeling method based on the stimulus sequence and the transfer learning techniques using cross-subject data. As a result, we achieved the highest ITR of 250 bpm under a minute of calibration, which has been shown to be comparable to the state-of-the-art SSVEP paradigms. In summary, our work significantly improved the cVEP performance under few-shot learning, which is expected to expand the practicality and usability of cVEP-BCIs.

[ { "created": "Mon, 20 Nov 2023 08:20:51 GMT", "version": "v1" } ]

2023-11-21

[ [ "Miao", "Yining", "" ], [ "Shi", "Nanlin", "" ], [ "Huang", "Changxing", "" ], [ "Song", "Yonghao", "" ], [ "Chen", "Xiaogang", "" ], [ "Wang", "Yijun", "" ], [ "Gao", "Xiaorong", "" ] ]

The ultimate goal of brain-computer interfaces (BCIs) based on visual modulation paradigms is to achieve high-speed performance without the burden of extensive calibration. Code-modulated visual evoked potential-based BCIs (cVEP-BCIs) modulated by broadband white noise (WN) offer various advantages, including increased communication speed, expanded encoding target capabilities, and enhanced coding flexibility. However, the complexity of the spatial-temporal patterns under broadband stimuli necessitates extensive calibration for effective target identification in cVEP-BCIs. Consequently, the information transfer rate (ITR) of cVEP-BCI under limited calibration usually stays around 100 bits per minute (bpm), significantly lagging behind state-of-the-art steady-state visual evoked potential-based BCIs (SSVEP-BCIs), which achieve rates above 200 bpm. To enhance the performance of cVEP-BCIs with minimal calibration, we devised an efficient calibration stage involving a brief single-target flickering, lasting less than a minute, to extract generalizable spatial-temporal patterns. Leveraging the calibration data, we developed two complementary methods to construct cVEP temporal patterns: the linear modeling method based on the stimulus sequence and the transfer learning techniques using cross-subject data. As a result, we achieved the highest ITR of 250 bpm under a minute of calibration, which has been shown to be comparable to the state-of-the-art SSVEP paradigms. In summary, our work significantly improved the cVEP performance under few-shot learning, which is expected to expand the practicality and usability of cVEP-BCIs.

2204.08970

Zhihao Li

Zhihao Li, Si Yi, Zhan Ma

Rendering Nighttime Image Via Cascaded Color and Brightness Compensation

Accepted by NTIRE 2022 (CVPR Workshop)

null

cs.CV eess.IV

http://creativecommons.org/licenses/by/4.0/

Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website: https://njuvision.github.io/CBUnet.

[ { "created": "Tue, 19 Apr 2022 16:15:31 GMT", "version": "v1" }, { "created": "Thu, 21 Apr 2022 17:23:11 GMT", "version": "v2" } ]

2022-04-22

[ [ "Li", "Zhihao", "" ], [ "Yi", "Si", "" ], [ "Ma", "Zhan", "" ] ]

Image signal processing (ISP) is crucial for camera imaging, and neural networks (NN) solutions are extensively deployed for daytime scenes. The lack of sufficient nighttime image dataset and insights on nighttime illumination characteristics poses a great challenge for high-quality rendering using existing NN ISPs. To tackle it, we first built a high-resolution nighttime RAW-RGB (NR2R) dataset with white balance and tone mapping annotated by expert professionals. Meanwhile, to best capture the characteristics of nighttime illumination light sources, we develop the CBUnet, a two-stage NN ISP to cascade the compensation of color and brightness attributes. Experiments show that our method has better visual quality compared to traditional ISP pipeline, and is ranked at the second place in the NTIRE 2022 Night Photography Rendering Challenge for two tracks by respective People's and Professional Photographer's choices. The code and relevant materials are avaiable on our website: https://njuvision.github.io/CBUnet.

1702.05724

Daniel M\'endez Fern\'andez

Marco Kuhrmann and Daniel M\'endez Fern\'andez and Thomas Ternit\'e

On the Use of Variability Operations in the V-Modell XT Software Process Line

Journal of Software: Evolution and Process, 2015

null

10.1002/smr.1751

null

cs.SE

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Software process lines provide a systematic approach to develop and manage software processes. It defines a reference process containing general process assets, whereas a well-defined customization approach allows process engineers to create new process variants, e.g., by extending or modifying process assets. Variability operations are an instrument to realize flexibility by explicitly declaring required modifications, which are applied to create a procedurally generated company-specific process. However, little is known about which variability operations are suitable in practice. In this article, we present a study on the feasibility of variability operations to support the development of software process lines in the context of the V-Modell XT. We analyze which variability operations are defined and practically used. We provide an initial catalog of variability operations as an improvement proposal for other process models. Our findings show that 69 variability operation types are defined across several metamodel versions of which, however, 25 remain unused. The found variability operations allow for systematically modifying the content of process model elements and the process documentation, and they allow for altering the structure of a process model and its description. Furthermore, we also find that variability operations can help process engineers to compensate process metamodel evolution.

[ { "created": "Sun, 19 Feb 2017 08:58:12 GMT", "version": "v1" } ]

2017-02-21

[ [ "Kuhrmann", "Marco", "" ], [ "Fernández", "Daniel Méndez", "" ], [ "Ternité", "Thomas", "" ] ]

Software process lines provide a systematic approach to develop and manage software processes. It defines a reference process containing general process assets, whereas a well-defined customization approach allows process engineers to create new process variants, e.g., by extending or modifying process assets. Variability operations are an instrument to realize flexibility by explicitly declaring required modifications, which are applied to create a procedurally generated company-specific process. However, little is known about which variability operations are suitable in practice. In this article, we present a study on the feasibility of variability operations to support the development of software process lines in the context of the V-Modell XT. We analyze which variability operations are defined and practically used. We provide an initial catalog of variability operations as an improvement proposal for other process models. Our findings show that 69 variability operation types are defined across several metamodel versions of which, however, 25 remain unused. The found variability operations allow for systematically modifying the content of process model elements and the process documentation, and they allow for altering the structure of a process model and its description. Furthermore, we also find that variability operations can help process engineers to compensate process metamodel evolution.

1711.11017

Ethan Perez

Simon Brodeur, Ethan Perez, Ankesh Anand, Florian Golemo, Luca Celotti, Florian Strub, Jean Rouat, Hugo Larochelle, Aaron Courville

HoME: a Household Multimodal Environment

Presented at NIPS 2017's Visually-Grounded Interaction and Language Workshop

null

cs.AI cs.CL cs.CV cs.RO cs.SD eess.AS

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

We introduce HoME: a Household Multimodal Environment for artificial agents to learn from vision, audio, semantics, physics, and interaction with objects and other agents, all within a realistic context. HoME integrates over 45,000 diverse 3D house layouts based on the SUNCG dataset, a scale which may facilitate learning, generalization, and transfer. HoME is an open-source, OpenAI Gym-compatible platform extensible to tasks in reinforcement learning, language grounding, sound-based navigation, robotics, multi-agent learning, and more. We hope HoME better enables artificial agents to learn as humans do: in an interactive, multimodal, and richly contextualized setting.

[ { "created": "Wed, 29 Nov 2017 18:45:59 GMT", "version": "v1" } ]

2017-11-30

[ [ "Brodeur", "Simon", "" ], [ "Perez", "Ethan", "" ], [ "Anand", "Ankesh", "" ], [ "Golemo", "Florian", "" ], [ "Celotti", "Luca", "" ], [ "Strub", "Florian", "" ], [ "Rouat", "Jean", "" ], [ "Larochelle", "Hugo", "" ], [ "Courville", "Aaron", "" ] ]

We introduce HoME: a Household Multimodal Environment for artificial agents to learn from vision, audio, semantics, physics, and interaction with objects and other agents, all within a realistic context. HoME integrates over 45,000 diverse 3D house layouts based on the SUNCG dataset, a scale which may facilitate learning, generalization, and transfer. HoME is an open-source, OpenAI Gym-compatible platform extensible to tasks in reinforcement learning, language grounding, sound-based navigation, robotics, multi-agent learning, and more. We hope HoME better enables artificial agents to learn as humans do: in an interactive, multimodal, and richly contextualized setting.

1709.03421

Riccardo Sven Risuleo

Riccardo Sven Risuleo, Giulio Bottegal, H{\aa}kan Hjalmarsson

Modeling and identification of uncertain-input systems

27 Pages, submitted to Automatica

null

cs.SY

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

In this work, we present a new class of models, called uncertain-input models, that allows us to treat system-identification problems in which a linear system is subject to a partially unknown input signal. To encode prior information about the input or the linear system, we use Gaussian-process models. We estimate the model from data using the empirical Bayes approach: the input and the impulse responses of the linear system are estimated using the posterior means of the Gaussian-process models given the data, and the hyperparameters that characterize the Gaussian-process models are estimated from the marginal likelihood of the data. We propose an iterative algorithm to find the hyperparameters that relies on the EM method and results in simple update steps. In the most general formulation, neither the marginal likelihood nor the posterior distribution of the unknowns is tractable. Therefore, we propose two approximation approaches, one based on Markov-chain Monte Carlo techniques and one based on variational Bayes approximation. We also show special model structures for which the distributions are treatable exactly. Through numerical simulations, we study the application of the uncertain-input model to the identification of Hammerstein systems and cascaded linear systems. As part of the contribution of the paper, we show that this model structure encompasses many classical problems in system identification such as classical PEM, Hammerstein models, errors-in-variables problems, blind system identification, and cascaded linear systems. This allows us to build a systematic procedure to apply the algorithms proposed in this work to a wide class of classical problems.

[ { "created": "Mon, 11 Sep 2017 14:53:38 GMT", "version": "v1" } ]

2017-09-12

[ [ "Risuleo", "Riccardo Sven", "" ], [ "Bottegal", "Giulio", "" ], [ "Hjalmarsson", "Håkan", "" ] ]

In this work, we present a new class of models, called uncertain-input models, that allows us to treat system-identification problems in which a linear system is subject to a partially unknown input signal. To encode prior information about the input or the linear system, we use Gaussian-process models. We estimate the model from data using the empirical Bayes approach: the input and the impulse responses of the linear system are estimated using the posterior means of the Gaussian-process models given the data, and the hyperparameters that characterize the Gaussian-process models are estimated from the marginal likelihood of the data. We propose an iterative algorithm to find the hyperparameters that relies on the EM method and results in simple update steps. In the most general formulation, neither the marginal likelihood nor the posterior distribution of the unknowns is tractable. Therefore, we propose two approximation approaches, one based on Markov-chain Monte Carlo techniques and one based on variational Bayes approximation. We also show special model structures for which the distributions are treatable exactly. Through numerical simulations, we study the application of the uncertain-input model to the identification of Hammerstein systems and cascaded linear systems. As part of the contribution of the paper, we show that this model structure encompasses many classical problems in system identification such as classical PEM, Hammerstein models, errors-in-variables problems, blind system identification, and cascaded linear systems. This allows us to build a systematic procedure to apply the algorithms proposed in this work to a wide class of classical problems.

2110.04678

Ankit Parag Shah

Rita Singh, Ankit Shah, Hira Dhamyal

An Overview of Techniques for Biomarker Discovery in Voice Signal

Last two authors contributed equally to the paper

null

cs.SD cs.AI eess.AS

http://creativecommons.org/licenses/by-nc-nd/4.0/

This paper reflects on the effect of several categories of medical conditions on human voice, focusing on those that may be hypothesized to have effects on voice, but for which the changes themselves may be subtle enough to have eluded observation in standard analytical examinations of the voice signal. It presents three categories of techniques that can potentially uncover such elusive biomarkers and allow them to be measured and used for predictive and diagnostic purposes. These approaches include proxy techniques, model-based analytical techniques and data-driven AI techniques.

[ { "created": "Sun, 10 Oct 2021 01:39:28 GMT", "version": "v1" } ]

2021-10-12

[ [ "Singh", "Rita", "" ], [ "Shah", "Ankit", "" ], [ "Dhamyal", "Hira", "" ] ]

This paper reflects on the effect of several categories of medical conditions on human voice, focusing on those that may be hypothesized to have effects on voice, but for which the changes themselves may be subtle enough to have eluded observation in standard analytical examinations of the voice signal. It presents three categories of techniques that can potentially uncover such elusive biomarkers and allow them to be measured and used for predictive and diagnostic purposes. These approaches include proxy techniques, model-based analytical techniques and data-driven AI techniques.

2308.01286

Diptapriyo Majumdar

Enumeration Kernels of Polynomial Size for Cuts of Bounded Degree

There have been major revision in the technicalities and proofs of the paper

null

cs.DS cs.DM

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Enumeration kernelization was first proposed by Creignou et al. [TOCS 2017] and was later refined by Golovach et al. [JCSS 2022] into two different variants: fully-polynomial enumeration kernelization and polynomial-delay enumeration kernelization. In this paper, we consider the DEGREE-d-CUT problem from the perspective of (polynomial-delay) enumeration kenrelization. Given an undirected graph G = (V, E), a cut F = (A, B) is a degree-d-cut of G if every $u \in A$ has at most d neighbors in B and every $v \in B$ has at most d neighbors in A. Checking the existence of a degree-d-cut in a graph is a well-known NP-hard problem and is well-studied in parameterized complexity [Algorithmica 2021, IWOCA 2021]. This problem also generalizes a well-studied problem MATCHING CUT (set d = 1) that has been a central problem in the literature of polynomial-delay enumeration kernelization. In this paper, we study three different enumeration variants of this problem, ENUM DEGREE-d-CUT, ENUM MIN-DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT that intends to enumerate all the d-cuts, all the minimal d-cuts and all the maximal degree-d-cuts respectively. We consider various structural parameters of the input and for every fixed $d \geq 1$, we provide polynomial-delay enumeration kernelizations of polynomial size for ENUM DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT and fully-polynomial enumeration kernels of polynomial size for ENUM MIN-DEGREE-d-CUT.

[ { "created": "Wed, 2 Aug 2023 17:18:19 GMT", "version": "v1" }, { "created": "Thu, 30 Nov 2023 09:14:39 GMT", "version": "v2" }, { "created": "Fri, 1 Dec 2023 07:32:28 GMT", "version": "v3" }, { "created": "Fri, 2 Feb 2024 18:15:47 GMT", "version": "v4" }, { "created": "Sat, 27 Apr 2024 17:42:26 GMT", "version": "v5" } ]

2024-04-30

[ [ "Majumdar", "Diptapriyo", "" ] ]

Enumeration kernelization was first proposed by Creignou et al. [TOCS 2017] and was later refined by Golovach et al. [JCSS 2022] into two different variants: fully-polynomial enumeration kernelization and polynomial-delay enumeration kernelization. In this paper, we consider the DEGREE-d-CUT problem from the perspective of (polynomial-delay) enumeration kenrelization. Given an undirected graph G = (V, E), a cut F = (A, B) is a degree-d-cut of G if every $u \in A$ has at most d neighbors in B and every $v \in B$ has at most d neighbors in A. Checking the existence of a degree-d-cut in a graph is a well-known NP-hard problem and is well-studied in parameterized complexity [Algorithmica 2021, IWOCA 2021]. This problem also generalizes a well-studied problem MATCHING CUT (set d = 1) that has been a central problem in the literature of polynomial-delay enumeration kernelization. In this paper, we study three different enumeration variants of this problem, ENUM DEGREE-d-CUT, ENUM MIN-DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT that intends to enumerate all the d-cuts, all the minimal d-cuts and all the maximal degree-d-cuts respectively. We consider various structural parameters of the input and for every fixed $d \geq 1$, we provide polynomial-delay enumeration kernelizations of polynomial size for ENUM DEGREE-d-CUT and ENUM MAX-DEGREE-d-CUT and fully-polynomial enumeration kernels of polynomial size for ENUM MIN-DEGREE-d-CUT.

0810.2529

Alireza Bayesteh

Jamshid Abouei, Alireza Bayesteh, Masoud Ebrahimi, and Amir K. Khandani

On the Throughput Maximization in Dencentralized Wireless Networks

Submitted to IEEE Transactions on Information Theory

null

cs.IT math.IT

http://arxiv.org/licenses/nonexclusive-distrib/1.0/

A distributed single-hop wireless network with $K$ links is considered, where the links are partitioned into a fixed number ($M$) of clusters each operating in a subchannel with bandwidth $\frac{W}{M}$. The subchannels are assumed to be orthogonal to each other. A general shadow-fading model, described by parameters $(\alpha,\varpi)$, is considered where $\alpha$ denotes the probability of shadowing and $\varpi$ ($\varpi \leq 1$) represents the average cross-link gains. The main goal of this paper is to find the maximum network throughput in the asymptotic regime of $K \to \infty$, which is achieved by: i) proposing a distributed and non-iterative power allocation strategy, where the objective of each user is to maximize its best estimate (based on its local information, i.e., direct channel gain) of the average network throughput, and ii) choosing the optimum value for $M$. In the first part of the paper, the network hroughput is defined as the \textit{average sum-rate} of the network, which is shown to scale as $\Theta (\log K)$. Moreover, it is proved that in the strong interference scenario, the optimum power allocation strategy for each user is a threshold-based on-off scheme. In the second part, the network throughput is defined as the \textit{guaranteed sum-rate}, when the outage probability approaches zero. In this scenario, it is demonstrated that the on-off power allocation scheme maximizes the throughput, which scales as $\frac{W}{\alpha \varpi} \log K$. Moreover, the optimum spectrum sharing for maximizing the average sum-rate and the guaranteed sum-rate is achieved at M=1.

[ { "created": "Tue, 14 Oct 2008 19:40:22 GMT", "version": "v1" } ]

2008-10-15

[ [ "Abouei", "Jamshid", "" ], [ "Bayesteh", "Alireza", "" ], [ "Ebrahimi", "Masoud", "" ], [ "Khandani", "Amir K.", "" ] ]

A distributed single-hop wireless network with $K$ links is considered, where the links are partitioned into a fixed number ($M$) of clusters each operating in a subchannel with bandwidth $\frac{W}{M}$. The subchannels are assumed to be orthogonal to each other. A general shadow-fading model, described by parameters $(\alpha,\varpi)$, is considered where $\alpha$ denotes the probability of shadowing and $\varpi$ ($\varpi \leq 1$) represents the average cross-link gains. The main goal of this paper is to find the maximum network throughput in the asymptotic regime of $K \to \infty$, which is achieved by: i) proposing a distributed and non-iterative power allocation strategy, where the objective of each user is to maximize its best estimate (based on its local information, i.e., direct channel gain) of the average network throughput, and ii) choosing the optimum value for $M$. In the first part of the paper, the network hroughput is defined as the \textit{average sum-rate} of the network, which is shown to scale as $\Theta (\log K)$. Moreover, it is proved that in the strong interference scenario, the optimum power allocation strategy for each user is a threshold-based on-off scheme. In the second part, the network throughput is defined as the \textit{guaranteed sum-rate}, when the outage probability approaches zero. In this scenario, it is demonstrated that the on-off power allocation scheme maximizes the throughput, which scales as $\frac{W}{\alpha \varpi} \log K$. Moreover, the optimum spectrum sharing for maximizing the average sum-rate and the guaranteed sum-rate is achieved at M=1.

2010.08660

Manas Gaur

Manas Gaur, Keyur Faldu, Amit Sheth

Semantics of the Black-Box: Can knowledge graphs help make deep learning systems more interpretable and explainable?

6 pages + references, 4 figures, Accepted to IEEE internet computing 2020

null

cs.AI

http://creativecommons.org/licenses/by/4.0/

The recent series of innovations in deep learning (DL) have shown enormous potential to impact individuals and society, both positively and negatively. The DL models utilizing massive computing power and enormous datasets have significantly outperformed prior historical benchmarks on increasingly difficult, well-defined research tasks across technology domains such as computer vision, natural language processing, signal processing, and human-computer interactions. However, the Black-Box nature of DL models and their over-reliance on massive amounts of data condensed into labels and dense representations poses challenges for interpretability and explainability of the system. Furthermore, DLs have not yet been proven in their ability to effectively utilize relevant domain knowledge and experience critical to human understanding. This aspect is missing in early data-focused approaches and necessitated knowledge-infused learning and other strategies to incorporate computational knowledge. This article demonstrates how knowledge, provided as a knowledge graph, is incorporated into DL methods using knowledge-infused learning, which is one of the strategies. We then discuss how this makes a fundamental difference in the interpretability and explainability of current approaches, and illustrate it with examples from natural language processing for healthcare and education applications.

[ { "created": "Fri, 16 Oct 2020 22:55:23 GMT", "version": "v1" }, { "created": "Sun, 1 Nov 2020 02:28:43 GMT", "version": "v2" }, { "created": "Tue, 3 Nov 2020 15:52:55 GMT", "version": "v3" }, { "created": "Fri, 11 Dec 2020 23:03:11 GMT", "version": "v4" } ]

2020-12-15

[ [ "Gaur", "Manas", "" ], [ "Faldu", "Keyur", "" ], [ "Sheth", "Amit", "" ] ]

The recent series of innovations in deep learning (DL) have shown enormous potential to impact individuals and society, both positively and negatively. The DL models utilizing massive computing power and enormous datasets have significantly outperformed prior historical benchmarks on increasingly difficult, well-defined research tasks across technology domains such as computer vision, natural language processing, signal processing, and human-computer interactions. However, the Black-Box nature of DL models and their over-reliance on massive amounts of data condensed into labels and dense representations poses challenges for interpretability and explainability of the system. Furthermore, DLs have not yet been proven in their ability to effectively utilize relevant domain knowledge and experience critical to human understanding. This aspect is missing in early data-focused approaches and necessitated knowledge-infused learning and other strategies to incorporate computational knowledge. This article demonstrates how knowledge, provided as a knowledge graph, is incorporated into DL methods using knowledge-infused learning, which is one of the strategies. We then discuss how this makes a fundamental difference in the interpretability and explainability of current approaches, and illustrate it with examples from natural language processing for healthcare and education applications.

Datasets:

LLMsForHepth
/

cs_primary_200k

Collection including LLMsForHepth/cs_primary_200k

Base datasets