Edgar Schönfeld

I get the same errors on a V100 on a cluster, using X11 forwarding. Did anybody make it work?

Hi, Thanks for your question. This repository is a copy of https://github.com/ajbrock/BigGAN-PyTorch, with modifications to turn BigGAN into U-NetGAN. Thus, you should have a look at the instructions in their...

Hi, Yes, we used the official LPIPS repository. There has been an API change (see https://github.com/richzhang/PerceptualSimilarity/issues/79), so use ``` # pip install lpips from lpips import LPIPS lpips_metric = LPIPS(net='vgg').cuda()...

Hi, I am assuming you refer to the ability of the model to apply different noise to different part of the image. For this, you can use the following function...

Hi, The problem was caused by the last commit in the updateEMA function: https://github.com/boschresearch/OASIS/commit/39fe99b7fa492acfa1b66ca64a17fd535f9c6b69 (using model.netEMA instead of model.module.netEMA). I fixed it now with a new commit.

The ```num_workers``` can be used as argument in the following function: https://github.com/boschresearch/OASIS/blob/9d4470e9d485cc43c305091d83c5d3f1a3221c5e/dataloaders/dataloaders.py#L23 To make it easier for the user, I now added an optional ```--num_workers``` flag in case you don't...

Hi, All layers defined in the generator's init function are trainable: https://github.com/boschresearch/OASIS/blob/8fc1cf06e67fdbb641760080099ea720128110fd/models/generator.py#L7-L23 Looking at the StyleGAN-NADA page (https://stylegan-nada.github.io/), I see that you need a ```G_train``` and ```G_frozen```. You can freeze...

Hi, it's very much possible that the hyperparameters have to be adjusted for each ablation, especially the early stopping step for the final classifier. That should be done on the...

Hi, Let's assume we have 2 trained encoders to extract latent features. The question now is "how is the classifier trained and tested?". Let's take the example of the CUB...

The attributes for each dataset can be found in the dropbox folder that is linked in the readme. That folder just has to be copied in the repository, and then...