Morphis is an advanced deepfake application that brings real-time face-swapping to video conferencing. Leveraging InsightFace for accurate face detection and swapping, it ensures a smooth, GPU-optimized experience, with CPU support when necessary. Built with PyQt5, Morphis offers a clean, intuitive interface, allowing users to toggle face-swapping effects seamlessly in their live video feed..
- Real-time Face Swapping: Use deepfake technology to swap faces in a live video feed.
- GPU Support: Automatically detects GPU availability and utilizes it for faster processing when possible.
- Intuitive UI: Built with PyQt5, offering easy controls for toggling face swap and GPU status.
- Default Face Detection: Preloads a default image for face swapping, with verification to ensure the image contains a detectable face.
We are actively working on adding more advanced features, including:
- Live Voice Cloning: In addition to face-swapping, future versions will support real-time voice cloning, allowing for more immersive deepfake experiences.
- Enhanced UI and Controls: Improved usability and additional settings for fine-tuning the deepfake output.
- Additional Model Support: Experimenting with different face and voice models to expand compatibility and enhance output quality.
- Python 3.x
- InsightFace
- PyQt5
- OpenCV
- ONNX Runtime (for GPU support)
Clone the repository and install the dependencies:
git clone https://github.com/yourusername/Morphis-Ultimate-DeepFake-Project.git
cd Morphis-Ultimate-DeepFake-Project
pip install -r requirements.txtThis project requires the inswapper_128.onnx model file for face swapping. Please download the model and place it in the following directory:
src/inswapper_128.onnx
You can customize the target face used for deepfake swapping by replacing the default image at:
assets/face.jpg