This is a high-definition video digital human project. First, I use Wav2Lip to modify the mouth shape, and then use CodeFormer for high-definition processing. This project has a better effect than Wav2Lip-GFPGAN, because CodeFormer performs better in facial restoration。

This is a local version of Wav2Lip-CodeFormer.

1. git clone this repository.
2. python 1-Setup.py

It will ask if you want to use an existing conda env or not.

https://github.com/langzizhixin/Wav2Lip-CodeFormer

https://github.com/sczhou/CodeFormer

https://github.com/Rudrabha/Wav2Lip