search

joonson / syncnet_python

Out of time: automated lip sync in the wild
MIT License
681 stars 149 forks source link

Full face or mouth region only #9

Closed ahmedelm closed 5 years ago

ahmedelm commented 5 years ago

Hi,

Is the model uploaded here trained on the full faces, or just the mouth region ? The paper says that the network takes the mouth region as input, but the testing script takes in the full face.

Thanks in advance

joonson commented 5 years ago

The model has been trained on the full faces, as in the example video.

ahmedelm commented 5 years ago

Many thanks for your response.

What about the model in https://github.com/joonson/yousaidthat ? Is it trained on the face aswell, or the lips ?

joonson commented 5 years ago

It is also trained on the full face.