BrainWi2ard is a user-friendly GUI software that integrates features from multiple deep learning projects to create a comprehensive tool for real-time voice cloning, deepfake video generation, and lip-syncing. This software is compatible with macOS, Linux, and Windows.
- Real-Time Voice Cloning: Clone a voice in 5 seconds to generate arbitrary speech in real-time.
- Deepfake Video Generation: Create deepfake videos using pre-trained models or train your own.
- Lip-Syncing: Accurately lip-sync videos in any language.
- Python 3.7+
- ffmpeg
- PyTorch
-
Clone the Repositories:
git clone https://github.com/iperov/DeepFaceLive.git git clone https://github.com/CorentinJ/Real-Time-Voice-Cloning.git git clone https://github.com/snehitvaddi/Deepfake-using-Wave2Lip.git git clone https://github.com/hectorgie/DeepFaceLab.git
-
Install Requirements:
cd DeepFaceLive pip install -r requirements.txt cd ../Real-Time-Voice-Cloning pip install -r requirements.txt cd ../Deepfake-using-Wave2Lip pip install -r requirements.txt cd ../DeepFaceLab pip install -r requirements.txt
-
Download Pre-trained Models (if applicable):
- Follow the instructions in each repository to download and set up pre-trained models.
-
Run the GUI:
- Implement a unified GUI using a framework like PyQt or Tkinter, integrating features from all repositories.
- Navigate to the Real-Time-Voice-Cloning directory:
cd Real-Time-Voice-Cloning
- Run the toolbox:
python demo_toolbox.py -d <datasets_root>
- Follow the instructions in the DeepFaceLab repository to create and train deepfake models.
- Use DeepFaceLive to apply real-time face swaps during video calls or streams.
- Navigate to the Deepfake-using-Wave2Lip directory:
cd Deepfake-using-Wave2Lip
- Run the lip-sync script with your base video and audio file:
python inference.py --checkpoint_path <path_to_checkpoint> --face <path_to_video> --audio <path_to_audio>
Contributions are welcome! Please fork this repository and submit pull requests.
This project is licensed under the MIT License.
-->