search

BadToBest / EchoMimic

Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://badtobest.github.io/echomimic.html
Apache License 2.0
2.28k stars 265 forks source link

Pose driven inference #110

Open alasokolova opened 1 month ago

alasokolova commented 1 month ago

Hi, how can i infer based on just pose, without audio at all (like face reenactment task)? Using all the landmarks, including mouth.

min-star commented 1 month ago

same problem