Can not reptoduce the face and eye movement.

[OrvilleQ]

Hello.

I am trying to reproduce the character's face as well as eye movements in the demo video you provided. Since the female hair in that video is floating all the time, I assumed that it was generated from a single still image. So I generated a similar image and expanded it to 240 batches and generated it based on the image examples in the repository. However, the generated video only has the mouth movement, not the character's face and eyes following the speech as in the demo video.

How can I achieve the effect of the demo video?

Thank you for your help.

[SamKhoze]

@OrvilleQ we are adding a new node that is based on SadTalker and you can generate video from an image, however, it only works with close-up faces, the node will be added shortly