TMElyralab / MuseTalk

MuseTalk: Real-Time High Quality Lip Synchorization with Latent Space Inpainting

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

different audio problem

TigerHH6866 opened this issue · comments

by using audio 1(the one target avatar speaking) to create avatar
then using audio 2(different people)
lipsync result seems to be different??
is it limited to use audio same as avatar created?

It should work in your case. Creating avatar only pre-save vision information. Could you share your results?

It should work in your case. Creating avatar only pre-save vision information. Could you share your results?

found that using different people audio get little different result. so i turn to bbox -30 close lip firstly.
then all comes same result