search

harlanhong / CVPR2022-DaGAN

Official code for CVPR2022 paper: Depth-Aware Generative Adversarial Network for Talking Head Video Generation
https://harlanhong.github.io/publications/dagan.html
Other
957 stars 125 forks source link

Full-body animation #60

Open cihankaradogan opened 1 year ago

cihankaradogan commented 1 year ago

I want to use your amazing work for full body animation. In an issue, it says something like it works if the same body parts exist in both the driving video and the source image. Does this mean that the current pretrained models will work for the full body? Or should I train my own model? I thought it would work for full body as well since it's depth aware. Am I right?

harlanhong commented 1 year ago

You may need to train a new depth estimation network to estimate the depth of the body.

cihankaradogan commented 1 year ago

Can I train your Face-Depth-Network for full-body or should I replace it with another network which is made for full-body

harlanhong commented 1 year ago

Sure, both ways are ok.

hacker009-sudo commented 10 months ago

@cihankaradogan , Are you able to train for full body?