BadToBest / EchoMimic

Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://badtobest.github.io/echomimic.html
Apache License 2.0
2.26k stars 263 forks source link

Support for Selected Landmark #96

Open 345ishaan opened 1 month ago

345ishaan commented 1 month ago

Hi, does the model supports the following use case:

1) given a list of pkl files storing the landmark 2) model selects the best landmark from the set provided in 1) given the audio.