Generate SadTalker digital human videos using SD WebUI

Its principle is as follows (I didn't understand it, I just copied it):

1. Install the extension https://github.com/OpenTalker/SadTalker.git
2. Restart WebUI
3. There is a Sadtalker option in the interface, click on this tab
4. Upload pictures and audio on the left, and click generate on the right to create a video.

The effect feels better than D-ID, and it doesn't cost anything (if the webUI is deployed locally). Those without SD can also consider using the Google Colab version: https://colab.research.google.com/github/Winfredy/SadTalker/blob/main/quick_demo.ipynb