Yassine Alouini

Thanks for the link @gau-nernst but it is a smaller dataset if I am not wrong.

Thanks for the feedback and the link @gau-nernst. Isn't the "real" dataset the 22k one? Anyway, I will give it a try with the smaller one once I have time.

@luca-martial Any chance you are starting to work on this for computer vision? I would be happy to research the state-of-the art about perturbation methods for binary classification for CV...

> > @luca-martial Any chance you are starting to work on this for computer vision? I would be happy to research the state-of-the art about perturbation methods for binary classification...

@astrojuanlu That sounds reasonable indeed and thanks for the additional library. :+1: We are discussing this internally with other colleagues at the moment and will let you know if we...

Indeed, watershed is probably the limiting part for now. I will have a look at a pytorch implementation and update with more details. Thanks for the quick feedback.

I have found this: https://paperswithcode.com/paper/deep-watershed-transform-for-instance. Not sure how useful it could be, I probably need some exploration to see if we can extract the watershed part of it. Here is...

The following requirements seem to work: ``` matplotlib tqdm numpy opencv-python torch torchvision ```