Basic usage flow
- Preparing Audio Files: We recommend using mono audio files in .wav or .mp3 format, with a sampling rate of 16kHz for best results.
- Run the main program::
python main.py --audio_path your_audio_file.wav - View Results: The program displays an animation on the screen or generates a video file
Advanced Function Operation
- Real-time input mode::
python main.py --live
Real-time audio input using a microphone - Output Video Save::
Add the -output parameter to specify the save path - parameterization: Frame rate, mouth sensitivity and other parameters can be adjusted as required
caveat
- Pre-trained models may need to be downloaded for the first run
- Complex audio may take longer to process
- A quiet recording environment is recommended for best results
This answer comes from the articleLiteAvatar: Audio-driven 2D portraits of real-time interactive digital people running at 30fps on the CPUThe































