README.md

metadata

title: Deepfake Detection System
emoji: 🔍
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 6.2.0
app_file: app.py
pinned: false
license: mit

🔍 Deepfake Detection System

An AI-powered system to detect deepfakes in images and videos using frequency analysis, facial artifact detection, and temporal inconsistencies.

🚀 Live Demo

Try the system here: [Your Hugging Face Space Link]

✨ Features

• Image Analysis: Detect deepfakes in single images
• Video Analysis: Analyze videos for temporal inconsistencies
• Real-time Webcam: Test with your webcam
• Batch Processing: Upload multiple files at once
• Detailed Reports: Get comprehensive analysis results

🛠️ How It Works

1. Frequency Domain Analysis: Detects unnatural patterns in frequency spectrum
2. Face Artifact Detection: Identifies inconsistencies in facial features
3. Temporal Analysis: Checks frame-to-frame consistency in videos
4. Ensemble Learning: Combines multiple models for higher accuracy

📊 Technical Details

• Models Used: EfficientNet-B4, Xception
• Detection Methods: Frequency analysis, facial landmark analysis, CNN predictions
• Framework: PyTorch, Gradio
• Deployment: Hugging Face Spaces

📁 File Structure