Yuanbo Xiangli

Hi, thanks for reporting this issue. I checked the code and found there is a mistake I made when refactoring the project. Here's the fix: https://github.com/kam1107/RealnessGAN/blob/287679f3df59aa09f30f94af272dc759e9cf6612/train.py#L185-L191 Now it should be...

I noticed that the result you show comes from a randomly sampled batch and indeed in this case not all of them look real. This issue is primarily caused by...

I just realised that there might be some problem in your training data. Please refer to this link to prepare CelebA for training: https://github.com/tkarras/progressive_growing_of_gans#preparing-datasets-for-training. It centres and aligns the samples,...

To evaluate the model trained on the provided dataset, you should use the test set. If you are trying to render a new trajectory, the easiest way is to create...

In our setting, we transformed the coordinate system into ENU (https://en.wikipedia.org/wiki/Local_tangent_plane_coordinates) with the origin set to the center building. If you don't want to use GES you can also mathematically...

Hi, the training time depends on how you partition the scales. For example, on 56Leonard scene, we use a 4-stage training. The training time on a single A100 card is...

30w iters is more than enough for stage0 to converge. You can stop stage training early when the curve tends to converge (or already converges).

What's the PSNR you got for each stage? You may try to increase learning rate to see if it helps. Also remember to load checkpoint from previous training stage when...

I think there's a mistake in render_path function. I manually assigned the stage variable to 0 in function render(..). Can you try to replace it with e.g. stage=3 to see...

Can you show some examples? My retrain results look fine 🤔