search

cleardusk / 3DDFA_V2

The official PyTorch implementation of Towards Fast, Accurate and Stable 3D Dense Face Alignment, ECCV 2020.
MIT License
2.92k stars 515 forks source link

Questions about implementing 3D Aided Short-video-synthesis. #117

Closed seven-sent closed 2 years ago

seven-sent commented 2 years ago

Thank you for sharing your wonderful work!

I would like to implement 3D Aided Short-video-synthesis based on your article, but I got some doubts. You noted in your article that you were using Face profiling for out-of-plane face moving. I have read the official Matlab code of Face Profiling. If I understand correctly, the out-of-plane face moving must use the 3D data (.mat) from the original image, right?

If there is only 2D images, is it possible to implement out-of-plane face moving?

Hope to hear from you. Best wishes.

cleardusk commented 2 years ago

You should fit an image to get the geometric infomation. You can refer to: https://github.com/zengxianyu/facepf

seven-sent commented 2 years ago

Thank you for your reply!