DeepFaceLive

Real-time face swapping and face animation application with GPU acceleration

🎭 What is DeepFaceLive?

DeepFaceLive is a powerful real-time face swapping and face animation application that allows you to:

• Face Swapping: Replace faces in real-time from webcam or video files using pre-trained models
• Face Animation: Control static face images using your webcam or video input
• Live Streaming: Use for video calls, streaming, and content creation
• GPU Acceleration: Optimized performance with NVIDIA CUDA and OpenCL support

Perfect for content creators, streamers, and anyone interested in real-time deepfake technology.

✨ Key Features

🔄 Real-time Face Swapping

• Swap your face with pre-trained celebrity models
• Support for webcam and video file input
• High-quality results with GPU acceleration
• Multiple face detection and alignment algorithms

🎬 Face Animation

• Animate static images using your facial expressions
• Real-time control with webcam input
• Fine-tuning parameters for optimal results
• 25+ FPS performance on modern GPUs

⚡ Performance Optimized

• GPU Acceleration: CUDA, TensorRT, and OpenCL support
• Multi-backend: ONNX Runtime and AVECL for optimal performance
• Real-time Processing: Low latency for live applications
• Memory Efficient: Optimized for various hardware configurations

🎯 Easy to Use

• Intuitive graphical interface
• Pre-trained models ready to use
• Comprehensive setup documentation
• Cross-platform support (Windows, Linux)

🚀 Quick Start

Prerequisites

• Python 3.8+ (Python 3.13 supported)
• NVIDIA GPU (recommended for best performance)
• CUDA 12.x and cuDNN 9.x (for GPU acceleration)
• 4GB+ RAM, 32GB+ paging file

Installation

1. Clone the repository

   git clone https://github.com/iperov/DeepFaceLive.git
   cd DeepFaceLive

2. Create virtual environment

   python -m venv venv
   # Windows
   venv\Scripts\activate
   # Linux/Mac
   source venv/bin/activate

3. Install dependencies

   # Full installation (recommended)
   pip install -r requirements.txt
   
   # Or minimal installation
   pip install -r requirements-minimal.txt

4. Run DeepFaceLive

   python main.py run DeepFaceLive --userdata-dir ./userdata

For detailed setup instructions, troubleshooting, and system requirements, see SETUP.md.

🎭 Available Face Models

DeepFaceLive comes with a variety of pre-trained face models ready to use:

Model	Preview	Examples
Keanu Reeves		View Examples
Bryan Greynolds		View Examples
Ewon Spice		View Examples
Emily Winston		View Examples

Note: These are AI-generated faces and do not represent real people. Any similarities are coincidental.

Custom Models

Want higher quality or better face matching? Train your own models using DeepFaceLab and import them into DeepFaceLive.

🎬 Face Animation Demo

Real-time face animation controlling static images with webcam input

🛠️ System Requirements

Minimum Requirements

• OS: Windows 10/11, Linux (Ubuntu 18.04+)
• CPU: Modern CPU with AVX instructions
• RAM: 4GB RAM, 32GB+ paging file
• GPU: DirectX12 compatible graphics card

Recommended Requirements

• GPU: NVIDIA RTX 2070+ / AMD Radeon RX 5700 XT+
• RAM: 16GB+ system RAM
• Storage: SSD with 10GB+ free space
• CUDA: Version 12.x with cuDNN 9.x

📖 Documentation

Setup Guides

• Complete Setup Guide - Detailed installation and configuration
• Windows Setup - Windows-specific instructions
• Linux Build Info - Linux compilation guide

Usage Guides

• Streaming Setup - Configure for live streaming
• Video Calls Setup - Setup for video conferencing
• Android Camera - Use phone as webcam

FAQ & Support

• User FAQ - Common questions and solutions
• Developer FAQ - Technical development info

🔧 Technical Architecture

DeepFaceLive uses a multi-backend architecture for optimal performance:

• ONNX Runtime: Face detection, marking, and swapping with CUDA/TensorRT support
• AVECL (OpenCL): Face merging with cross-platform GPU acceleration
• PyQt6: Modern, responsive user interface
• Multi-processing: Efficient pipeline processing for real-time performance

💻 Development

Project Structure

DeepFaceLive/
├── apps/DeepFaceLive/     # Main application
├── xlib/                  # Core libraries
├── modelhub/             # Pre-trained models
├── userdata/             # User settings and data
├── requirements.txt      # Full dependencies
├── requirements-minimal.txt  # Essential dependencies
└── SETUP.md             # Setup documentation

Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

📦 Downloads

Pre-built Releases

• Windows x64 (Yandex) - Standalone portable version
• Windows x64 (MEGA) - Alternative download

Portable versions include all dependencies and require no installation beyond GPU drivers.

🌐 Community

Official Channels

• Discord - Official community (English/Russian)
• GitHub Issues - Bug reports and feature requests

Community Forums

• MrDeepFakes - Largest English deepfake community
• DFLData.xyz - Chinese community forum
• QQ Group: 124500433 - Chinese community chat

🤝 Support the Project

How to Help

• ⭐ Star this repository on GitHub
• 🎭 Train and share high-quality face models
• 🐛 Report bugs and suggest improvements
• 📖 Improve documentation and tutorials

Donations

• Yoomoney - Russian payment system
• Bitcoin: bc1qewl062v70rszulml3f0mjdjrys8uxdydw3v6rq

⚖️ Legal Notice

This software is for educational and research purposes. Users are responsible for complying with local laws and regulations. The developers do not condone malicious use of this technology.

📄 License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.

---

Made with ❤️ by the DeepFaceLive community

Website • Documentation • Community • Issues