Fernando Pérez-García

I agree, I came here to open the same issue. I can give it a try and add some examples to thank the thousand times I've opened the website when...

> The thing with gpu support, is that it will not be available in pytorch DataLoader since they are cpu only . (correct ?) Oh yeah I've heard that. If...

> For instance for the reverse problem (#299). Since you just evaluate your model with GPU and you want to spatially inverse the prediction, you will be happier with a...

If we ever start tackling this, here are a couple of sources that might be useful: 1. [PyTorch forum – Affine transformation matrix parameters conversion](https://discuss.pytorch.org/t/affine-transformation-matrix-paramters-conversion/19522/18?u=fepegar) 1. [StackOverflow – Generating pytorch's...

Hey, @justusschock. Can you provide a more concrete use case? When you iterate over a dataset, images are loaded in the copied subject, so the subjects in the dataset shouldn't...

Is `images` a `SubjectsDataset`?

Hi, @ivezakis. You are using one process to load 8 images, so it will be 8 times slower. This is expected. To make it faster, you should use a `num_workers`...

@sarthakpati, according to the discussion in #301, it's not clear that this implementation would generally be preferred, and I believe it's now in a prototyping state. You can always check...

Hi, @Spenhouet. I think I understand the problem. It makes sense. Maybe we can add a flag to those transforms to not check for spatial consistency, as long as users...

This can be trivially solved for `RandomAffine`, but unfortunately not for `RandomElasticDeformation`. I can open a PR right away for the former, but I currently do not have the bandwidth...