Creating output for a custom face

grey-eye / talking-heads

Our implementation of "Few-Shot Adversarial Learning of Realistic Neural Talking Head Models" (Egor Zakharov et al.)

GNU General Public License v3.0

593 stars 110 forks source link

Creating output for a custom face #34

Closed sleebapaul closed 5 years ago

sleebapaul commented 5 years ago

Hi,

It could be a completely stupid question, but can I generate the video of the moving face for a custom image? All I'm left with is an image, not any other dataset related to it. Or it is specific to only the VoxCeleb2 dataset?

castelo-software commented 5 years ago

@sleebapaul, I haven't gotten around to making the fine-tuning phase work yet. You would need fine tuning to generate good quality frames for your custom image. Of course, this model only generates images. Once fine-tuning works, you'll still need to generate all frames and then combine them into a video.