An alternative project EchoMimic is open-sourced

tencent-ailab / V-Express

V-Express aims to generate a talking head video under the control of a reference image, an audio, and a sequence of V-Kps images.

2.18k stars 272 forks source link

Open lymhust opened 2 months ago

lymhust commented 2 months ago

EchoMimic is capable of generating portrait videos not only by audios and facial landmarks individually, but also by a combination of both audios and selected facial landmarks. Project link: https://badtobest.github.io/echomimic.html GitHub link: https://github.com/BadToBest/EchoMimic

FurkanGozukara commented 2 months ago

LivePortrait looks better than this have you compared?

Arvrairobo commented 1 month ago

@FurkanGozukara but how to generate audio driven image portraits with liveportraits? right now its only video driven? any thought on that?

FurkanGozukara commented 1 month ago

@Arvrairobo currently liveportait doesn't support audio driven sadly :/