search

sihyun-yu / digan

Official PyTorch implementation of Generating Videos with Dynamics-aware Implicit Generative Adversarial Networks (ICLR 2022).
https://sihyun.me/digan/
182 stars 19 forks source link

modulated_conv2d in ToRGBLayer #6

Closed zacjiang closed 2 years ago

zacjiang commented 2 years ago

Hi,

Thank you for your work.

I was looking at your code and I noticed that in ToRGBLayer, the modulated_conv2d function is used to generate the RGB frames. Does this mean that the network is not fully implicit but contains convolutions in the last layer, or did I miss something?

https://github.com/sihyun-yu/digan/blob/8368d5bfd73db38a9593ce10bb246cd7c37ddf9f/src/training/networks.py#L386

Thank you for your help!

sihyun-yu commented 2 years ago

It is fully implicit. Note that the kernel size is set to 1, which is equivalent to the linear layer.

zacjiang commented 2 years ago

Thank you!