Chieh Hubert Lin

Thanks for open-sourcing the great work! I wonder if your model supports inference with disparities given by users? In my scenario, I am interested in the results if I manually...

Awkward flickering while using opengl_viewer

2

Thanks for sharing your code! I have built the environment without any issue, and the results generated from cuda_renderer looks perfect to me. However, when I turn to use opengl_viewer,...

Why mIoU uses sum(intersect)/sum(union) instead of mean over IoU?

1

I found the mIoU evaluation codes here: https://github.com/CSAILVision/semantic-segmentation-pytorch/blob/8f27c9b97d2ca7c6e05333d5766d144bf7d8c31b/eval.py#L98 I am wondering what is the motivation of using `sum(intersect)/sum(union)` instead of the mean over sample-wise `intersect / union`? The former one...

Questions regarding the KID score implementation

1

Hi, Thanks for maintaining the library for GAN research! I noticed that the [KID implementation in this repository](https://github.com/GaParmar/clean-fid/blob/main/cleanfid/fid.py#L75-L77) is quite different from [the official KID implementation](https://github.com/mbinkowski/MMD-GAN/blob/678bb5e2d5f7b0bb8dd5c3591d7759e1bb3f8018/gan/compute_scores.py#L231) Could the author explain...

Reporting a minor bug

1

Hi, thanks for making the codes public! I found a minor bug [here](https://github.com/VITA-Group/TransGAN/blob/ff5aac42ed244b0a803baecb3af340ffeec53dc0/models_search/Celeba256_gen.py#L345). The variable `self.pos_embed` keeps the CPU version of the positional embedding. This is the root cause of...

I get zero accuracies by running the instructed command: ``` # coordconv version python train.py --arch coordconv_rendering -mb 16 -E 100 -L 0.005 --opt adam --l2 0.001 -mul 1 --use_mse_loss...

Questions regarding some design choices

3

Really impressive work and high-quality code release! I found several intriguing design choices while digging into the codebase, and looking for some clarifications or explanations of them: 1. **Blur in...

***TL;DR*** The author has modified the PGGAN architecture in their project, so your model pretrained on the original PGGAN codebase cannot be used here. You can delete it now. ***Related...

Two issues regarding the implementation

3

Thanks for your interest in COCO-GAN and your effort in implementing it in Pytorch! I took a quick look and found two issues (not sure if there is still any...

Worship!!

1

HOLY SHIT!!! THIS IS SO FUCKING BRILLIANT!!!!!!