Code availability update

elliottwu / unsup3d

(CVPR'20 Oral) Unsupervised Learning of Probably Symmetric Deformable 3D Objects from Images in the Wild

MIT License

1.19k stars 194 forks source link

Code availability update #1

Closed bc closed 4 years ago

bc commented 4 years ago

Hi! I would love to try this out and see what I can adapt it into. When do you target releasing the code?

elliottwu commented 4 years ago

Hi! Sorry for the delay! The initial version is now released.

bc commented 4 years ago

Awesome thank you! I'm installing it now on my mac through docker. Roughly, what was the amount of computation time it took per frame of input on the pretrained model?

elliottwu commented 4 years ago

The inference is realtime (25+fps) on CPU. The neural_renderer does not have CPU implementation, so the training/testing code will require GPU. I plan to release some demo code without using the neural_renderer in a few days, which will be able to run on CPU.