Skip to content

Timbered-manse640/image-reality-check

Repository files navigation

Image Reality Check

Version Manifest V3 License Privacy

Chrome extension that automatically detects AI-generated images, beauty filters, and photo manipulation — all offline, all client-side.

🇨🇳 中文文档


Features

  • AI-Generated Image Detection — 6-dimension analysis (frequency spectrum, noise, color, texture, symmetry, edges)
  • Error Level Analysis (ELA) — Colored heatmap revealing tampered or edited regions
  • Face Detection — BlazeFace-powered face detection + skin smoothness analysis
  • Beauty Filter Detection — Detects skin smoothing filters via texture variance
  • Screenshot Detection — Automatically skips screenshots and non-photo images
  • Auto-Detection — Scans all images on any webpage with progress indicator
  • Stamp Results — Marks suspicious images directly on the page (AI / Beauty Filter / Edited)
  • Face-Only Mode — Only analyze images containing faces (default ON)
  • Right-Click Analysis — Deep analysis of any image via context menu
  • Settings Toggle — Enable/disable auto-detection and face-only mode

Privacy

All processing happens locally in your browser. No images are uploaded. No data is sent to any server. Zero network requests. Your images never leave your device.

Installation

Developer Mode

  1. Download the latest release and unzip
  2. Open chrome://extensions/ in Chrome
  3. Enable Developer mode (top right)
  4. Click Load unpacked and select the unzipped directory

Chrome Web Store

Coming soon

How It Works

ELA (Error Level Analysis)

Re-compresses the image at a known quality level and compares the difference. Edited regions show higher error levels, visualized as a colored heatmap.

Skin Smoothness Detection

Uses BlazeFace to detect faces, then analyzes skin texture variance in face regions (forehead, cheeks, chin). Unnaturally low variance indicates beauty filter usage.

AI Generation Detection

Scores images across 6 dimensions:

  • Frequency Spectrum — AI images have steeper high-frequency falloff
  • Noise Uniformity — AI images have unnaturally uniform noise distribution
  • Color Distribution — AI images have smoother color histograms
  • Texture Regularity — AI images have more regular LBP texture patterns
  • Facial Symmetry — AI-generated faces tend to be more symmetrical
  • Edge Coherence — Detects AI-typical edge artifacts

Screenshot Detection

Identifies screenshots via flat color block ratio, color diversity, and noise level. Screenshots are automatically skipped in auto-detection mode.

Tech Stack

  • Manifest V3 — Modern Chrome extension architecture
  • TensorFlow.js — Client-side ML inference
  • BlazeFace — Real-time face detection model
  • Canvas API — Image pixel manipulation and analysis
  • Pure JavaScript — No build step, no frameworks

Contributing

Contributions are welcome!

  1. Fork the repository
  2. Create a feature branch (git checkout -b feature/amazing-feature)
  3. Commit your changes (git commit -m 'Add amazing feature')
  4. Push to the branch (git push origin feature/amazing-feature)
  5. Open a Pull Request

License

MIT License © 2026