LiveFacer is an advanced, high-performance, real-time face swapping application. Designed for seamless live video streaming and interactive use, it leverages state-of-the-art AI models to map and swap faces with incredible accuracy, high framerates, and minimal latency.
Whether you're a content creator, a developer, or just want to have fun on your next video call, LiveFacer delivers ultra-smooth performance tailored for modern GPUs.
We believe in open-source, and all the code is freely available here for anyone to compile. However, setting up AI environments, installing CUDA toolkits, configuring Python paths, and compiling from source can be a massive headache that takes hours.
Want to skip the setup and jump straight into the action?
We've prepared a Pre-compiled, 1-Click Portable Version that works out of the box! No Python installation, no environment headaches, no compiling errors. Just extract and double-click.
By purchasing the pre-compiled version, you also directly support the continuous development of this open-source project. ❤️
👉 Get the 1-Click Portable Version on Gumroad
- Real-Time Performance: Engineered for live camera feeds and high-FPS rendering.
- Hardware Acceleration: Native support for NVIDIA GPUs (CUDA), AMD (DirectML), and CPU fallbacks.
- Advanced Masking: Dynamic generation for Face, Eyes, Eyebrows, and Mouth masks using Convex Hull rendering to handle extreme facial expressions (like sticking your tongue out) without glitching.
- Live UI Control: Adjust swapping parameters, opacity, and mask boundaries on the fly.
- Multi-Face Mapping: Swap multiple faces simultaneously or target specific faces in a crowded frame.
- Color Correction: Built-in Poisson blending and color fixers to eliminate "blueish" or mismatched skin tones.
If you are a developer and prefer to build the project from scratch, LiveFacer is 100% open-source.
- Python 3.10+
- NVIDIA CUDA Toolkit & cuDNN (if using NVIDIA GPU)
- Visual Studio Build Tools (for C++ dependencies)
- FFmpeg
-
Clone the repository:
git clone https://github.com/ALPERDURUKAN/livefacer.git cd livefacer/code -
Install requirements:
pip install -r requirements.txt
-
Run the application:
python run.py --execution-provider cuda
(Note: Depending on your hardware, you may need to configure specific ONNX Runtime providers manually.)
Contributions, issues, and feature requests are welcome! Feel free to check the issues page.
This project is open-source and available under the standard MIT License. The core technology relies on several open-source AI models; please ensure you respect their respective licenses and ethical usage guidelines.
Disclaimer: LiveFacer is created for ethical, educational, and entertainment purposes only. Do not use this software to deceive others, create non-consensual content, or violate any laws.