better layout

README.md

@@ -4,7 +4,7 @@ emoji: 👀 😵‍💫 👀
 colorFrom: green
 colorTo: black
 sdk: gradio
-sdk_version: 2.9.4
+sdk_version: 2.9b22
 app_file: app.py
 pinned: false
 ---

app.py

@@ -5,10 +5,9 @@ import torch
 import numpy as np
 from PIL import Image, ImageOps
 from pathlib import Path
-import os
 import glob
 from autostereogram.sirds_converter import SirdsConverter
-from skimage import color
+from datetime import datetime
 
 feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
 model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large")
@@ -46,28 +45,75 @@ def process_image(image_path):
     stereo_image = stereo_converter.convert_depth_to_stereogram_with_sird(
         depth_image_padded, False, 0.5).astype(np.uint8)
 
-    return [depth_image_padded, stereo_image]
+    stereo_image_pil = Image.fromarray(stereo_image).convert('RGB')
+    image_name = f'stereo_image_{datetime.now().strftime("%Y%m%d_%H%M%S")}.jpg'
+    stereo_image_pil.save(image_name)
+    return [depth_image_padded, stereo_image, image_name]
 
 
-title = "Demo: zero-shot depth estimation with DPT + 3D Voxels reconstruction"
-description = "This demo is a variation from the original <a href='https://huggingface.co/spaces/nielsr/dpt-depth-estimation' target='_blank'>DPT Demo</a>. It uses the DPT model to predict the depth of an image and then reconstruct the 3D model as voxels."
+examples_images = [[f] for f in sorted(glob.glob('examples/*.jpg'))]
 
-examples = sorted(glob.glob('examples/*.jpg'))
+blocks = gr.Blocks()
 
-iface = gr.Interface(fn=process_image,
-                     inputs=[
-                         gr.inputs.Image(
-                             type="filepath", label="Input Image")
-                     ],
-                     outputs=[
-                         gr.outputs.Image(label="Predicted Depth", type="pil"),
-                         gr.outputs.Image(label="Stereogram", type="pil")
-                     ],
-                     description=description,
-                     examples=examples,
-                     allow_flagging="never",
-                     #  cache_examples=False
-                     )
+input_image = gr.Image(type="filepath", label="Input Image")
+predicted_depth = gr.Image(label="Predicted Depth", type="pil")
+autostereogram = gr.Image(label="Autostereogram", type="pil")
+file_download = gr.File(label="Download Image")
+
+
+def load_example(example_id):
+    processed_examples = [
+        component.preprocess_example(sample)
+        for component, sample in zip(
+            [input_image], examples_images[example_id]
+        )
+    ]
+    if len(processed_examples) == 1:
+        return processed_examples[0]
+    else:
+        return processed_examples
+
+
+with blocks:
+    gr.Markdown('''
+## Depth Image to Autostereogram (Magic Eye)
+This demo is a variation from the original [DPT Demo](https://huggingface.co/spaces/nielsr/dpt-depth-estimation).
+Zero-shot depth estimation from an image, then it uses [pystereogram](https://github.com/yxiao1996/pystereogram)
+to generate the autostereogram (Magic Eye)
+<base target="_blank">
+
+''')
+
+    with gr.Row():
+        examples_c = gr.components.Dataset(
+            components=[input_image],
+            samples=examples_images,
+            type="index",
+        )
+
+        examples_c._click_no_postprocess(
+            load_example,
+            inputs=[examples_c],
+            outputs=[input_image])
+
+    with gr.Row():
+        with gr.Column():
+            button = gr.Button("Predict")
+            button.click(fn=process_image, inputs=[input_image],
+                         outputs=[predicted_depth,
+                                  autostereogram, file_download],
+                         )
+
+    with gr.Row():
+        with gr.Column():
+            input_image.render()
+        with gr.Column():
+            predicted_depth.render()
+    with gr.Row():
+        autostereogram.render()
+    with gr.Row():
+        with gr.Column():
+            file_download.render()
 
 if __name__ == "__main__":
-    iface.launch(debug=True, enable_queue=False)
+    blocks.launch(debug=True)

requirements.txt

@@ -2,7 +2,7 @@ torch
 git+https://github.com/nielsrogge/transformers.git@add_dpt_redesign#egg=transformers
 numpy
 Pillow
-gradio==2.9.4
+gradio==2.9b22
 jinja2
 transformers
 scikit-image