Deepfake Detection Web Application

A web application that uses AI to detect deepfake images and videos. The application allows users to upload media, which is then analyzed to determine if it's real or fake.

Features

• Upload and analyze images and videos
• Separate sections for image and video detection
• Store media and analysis results in a database
• View detection history
• Get detailed analysis results with confidence scores

Architecture

The application is built using:

• Backend: Django with Django REST Framework
• Frontend: React.js
• Database: SQLite (default, configurable)
• AI Model: XceptionNet for deepfake detection

Project Structure

deepfake_detection/         # Django project root
│
├── deepfake_detection/     # Django settings and main URLs
│   ├── settings.py
│   ├── urls.py
│   └── ...
│
├── detector/               # Django app for deepfake detection
│   ├── models.py           # Database models for media and results
│   ├── views.py            # API views
│   ├── serializers.py      # DRF serializers
│   ├── urls.py             # App routes
│   ├── ai_detector.py      # Deepfake detection implementation
│   └── ...
│
├── media/                  # Uploaded media storage
│   ├── images/             # Uploaded images
│   └── videos/             # Uploaded videos
│
├── models/                 # AI model storage
│   └── xception_deepfake.h5  # XceptionNet model file
│
└── frontend/               # Frontend code
    ├── index.html          # Main HTML page
    └── api.js              # API integration code

How to Run

Backend Setup

1. Install the required Python packages:

   pip install -r requirements.txt
   

2. Navigate to the Django project directory:

   cd deepfake_detection
   

3. Download the placeholder model (or use a pre-trained XceptionNet model):

   python download_model.py
   

4. Run database migrations:

   python manage.py migrate
   

5. Start the Django development server:

   python manage.py runserver
   

The backend API will be available at: http://127.0.0.1:8000/api/

Frontend Setup

Since the frontend is built with plain HTML, CSS, and JavaScript with React loaded from CDN, you can simply open the frontend/index.html file in your browser.

For a better development experience, you can use a simple HTTP server:

cd frontend
python -m http.server

Then access the frontend at: http://localhost:8000

AI Model Information

This application uses XceptionNet for deepfake detection, which has shown excellent performance in detecting manipulated media:

• High accuracy (95%+ on benchmark datasets)
• Effective at detecting GAN-generated faces
• Good balance of performance and computational requirements
• Trained on diverse deepfake datasets

For detailed information about the AI model integration, refer to AI_MODEL_README.md.

API Endpoints

• GET /api/media/ - List all media items
• POST /api/media/ - Upload a new media item (image or video)
• GET /api/media/<id>/ - Get a specific media item with its detection result
• POST /api/media/<id>/analyze/ - Manually trigger analysis for a media item

Future Improvements

• Improve model accuracy with better training data
• Add user authentication
• Improve UI/UX
• Add more detailed analysis results
• Implement background processing for video analysis
• Add support for URL-based media analysis