FaceFormer trainning problem

The result video has mouth-only motion when I use pre-trained model. But when I use the model trained by myself, the result has whole-face motion. (vocaset)

Jun 07 '22 03:06 wangjio

请问你是用的自己的数据训练，还是用的voca的数据训练的？

Jun 10 '22 06:06 ZerRui

I also want to get the whole-face motion, but when I use the model trained by my self(vocaset), the result has mouth-only motion, how do you train it?

Jun 13 '22 02:06 fastcode3d

The result video has mouth-only motion when I use pre-trained model. But when I use the model trained by myself, the result has whole-face motion. (vocaset)

is there only output video ? and the fps is？

Jun 24 '22 03:06 foocker

Hey @wangjio , Did you able to train the model by yourself? Please refer the issue i'm getting while training. https://github.com/EvelynFan/FaceFormer/issues/40 Any suggestions are welcome.

Sep 23 '22 10:09 ujjawalcse

The result video has mouth-only motion when I use pre-trained model. But when I use the model trained by myself, the result has whole-face motion. (vocaset)

Can u share the pretrained model, it seems this repo close the shared pretrained model.

Jul 19 '23 06:07 JaggerZr