search

tencent-ailab / V-Express

V-Express aims to generate a talking head video under the control of a reference image, an audio, and a sequence of V-Kps images.
2.03k stars 250 forks source link

about the gt_generate_compare.mp4 #21

Closed NinoNeumann closed 1 month ago

NinoNeumann commented 1 month ago

Is this gt_generate_compare.mp4 the result of overfitting?

tiankuan93 commented 1 month ago

The training process is about learning a mapping relationship so that you can get the desired result when inference. If the model fits the training data well and also gets good results in the test data, this is exactly what we want the model to do.

Furthermore, if you look carefully at the area around the face in the video, you will see that there is still a lot of "generated" content in the generated video.