README.md

metadata

title: Deepfake Detection Bypass Utility
emoji: 👁
colorFrom: red
colorTo: yellow
sdk: gradio
sdk_version: 5.35.0
app_file: app_gradio.py
pinned: true
short_description: DF detection circumvention utility refactored for HF Spaces.

Image Detection Bypass Utility [Gradio Refactor]

Refactored Gradio app for HF Spaces of Image Detection Bypass, self-proclaimed app for the "Circumvention of AI Detection."

Strictly for research purposes -- please refer to the original repository for more in-depth documentation and up-to-date code.

This refactored repository will not be maintained.

---

Screenshot

---

Parameters / Controls → args mapping

When you click Run, the GUI builds a lightweight argument namespace (similar to a SimpleNamespace) and passes it to the worker. Below are the important mappings used by the GUI (so you know what your process_image should expect):

• args.noise_std — Gaussian noise STD (fraction of 255)
• args.clahe_clip — CLAHE clip limit
• args.tile — CLAHE tile size
• args.cutoff — Fourier cutoff (0.01–1.0)
• args.fstrength — Fourier strength (0–1)
• args.phase_perturb — phase perturbation STD (radians)
• args.randomness — Fourier randomness factor
• args.perturb — small pixel perturbations
• args.fft_mode — one of auto, ref, model
• args.fft_alpha — alpha exponent for 1/f model (used when fft_mode=='model')
• args.radial_smooth — radial smoothing bins for spectrum matching
• args.jpeg_cycles — number of lossy JPEG encode/decode cycles (camera sim)
• args.jpeg_qmin, args.jpeg_qmax — JPEG quality range used by camera sim
• args.vignette_strength — vignette intensity (0–1)
• args.chroma_strength — chromatic aberration strength (pixels)
• args.iso_scale — exposure multiplier (camera sim)
• args.read_noise — read noise in DN (camera sim)
• args.hot_pixel_prob — probability of hot pixels (camera sim)
• args.banding_strength — banding strength
• args.motion_blur_kernel — motion blur kernel size
• args.seed — integer seed or None when seed==0 in UI
• args.sim_camera — bool: run camera simulation path
• args.no_no_bayer — toggles Bayer/demosaic (True = enable RGGB demosaic)
• args.fft_ref — path to reference image (string) or None
• args.ref— path to auto white-balance reference image (string) or None

Tip: Your process_image(inpath, outpath, args) should be tolerant of missing keys (use getattr(args, 'name', default)), or accept the same SimpleNamespace object the GUI builds.

---

Gradio / Hugging Face Spaces

This repository includes a lightweight Gradio front-end (app_gradio.py) that wraps the existing process_image(inpath, outpath, args) pipeline. The Gradio app is suitable for local testing and for deployment to Hugging Face Spaces (Gradio-backed web apps).

Quick local run

1. Install dependencies:

pip install -r requirements.txt

2. Launch the Gradio app:

python3 app_gradio.py

Open http://localhost:7860 in your browser. The UI saves the uploaded image to a temporary file, calls the existing process_image pipeline, and returns the processed image.

Deploying to Hugging Face Spaces

1. Ensure the following are present at the repository root:

   • app_gradio.py (the Gradio entrypoint)
   • requirements.txt (must include gradio and any other runtime deps)

2. Push the repository to a new Space on Hugging Face (create a new Space and connect this repo or push to the Space's Git remote). Spaces will automatically run the Gradio app.

Notes & tips for Spaces:

• Keep default upload/processing sizes modest to avoid long CPU usage in the free tier.
• If your pipeline uses optional packages (OpenCV, piexif, etc.), make sure they are listed in requirements.txt so Spaces installs them.
• If processing is slow, consider reducing default image size or exposing fewer parameters to the main UI and keeping advanced controls hidden in an "Advanced" section.

Troubleshooting

• If Gradio is not installed, app_gradio.py will raise an error; add gradio to requirements.txt.
• Any import errors from image_postprocess will surface when calling the app; run the smoke test (python3 test_smoke_gradio.py) locally to validate imports and pipeline execution before pushing.

---

License

MIT — free to use and adapt. Please include attribution if you fork or republish.

---