Updated uflow

[mtailanian]

📝 Description

• Provide a clear summary of the changes and the issue that has been addressed.
• 🛠️ Fixes # (issue number)

fixed loss computation

✨ Changes

• added new anomaly segmentation method, from the uflow paper
• updated readme with final results, images, and paper link and citation

Select what type of change your PR is:

• 🐞 Bug fix (non-breaking change which fixes an issue)
• 🔨 Refactor (non-breaking change which refactors the code base)
• 🚀 New feature (non-breaking change which adds functionality)
• 💥 Breaking change (fix or feature that would cause existing functionality to not work as expected)
• 📚 Documentation update
• 🔒 Security update

✅ Checklist

Before you submit your pull request, please make sure you have completed the following steps:

• 📋 I have summarized my changes in the CHANGELOG and followed the guidelines for my type of change (skip for minor changes, documentation updates, and test enhancements).
• 📚 I have made the necessary updates to the documentation (if applicable).
• 🧪 I have written tests that support my changes and prove that my fix is effective or my feature works (if applicable).

For more information about code review checklists, see the Code Review Checklist.

[mtailanian]

Hello guys, I just wanted to ask: am I supposed to add something or correct something? should I update the branch?

[ashwinvaidya17]

Thanks for updating this. I have a few minor comments.

[mtailanian]

I guess it should be fine now... Should I update the branch in order to merge this?

---

Btw, I have one more question.
As you might have seen in the code, I included a method for finding an automatic segmentation of the anomalies, as this method is part of the paper. But it is unused in the current code.

Is there a good way to include it as part of the default outputs?
I saw that all the segmentations are actually obtained by maximizing some metric in testing sets, so I'm not sure if it's a good idea to try to add this method there... I guess not, but I can't think of any other place.

Or maybe it is just fine as it is... The code is there in case anyone wants to use it...

Any thoughts are appreciated.
Thanks!

[mtailanian]

Hi guys, hope everything is well.
Is there anything I could do to get this merged?
Thanks again!

[ashwinvaidya17]

@mtailanian the PR looks good to me. Do you have access to the logs here: https://github.com/openvinotoolkit/anomalib/actions/runs/9483018344/job/26129442354?pr=2104? If so, can you fix the pre-merge checks. Ideally, you should be able to run these locally as well. Just do pip install pre_commit and then run pre-commit run --all-files

[ashwinvaidya17]

I guess it should be fine now... Should I update the branch in order to merge this?

Btw, I have one more question. As you might have seen in the code, I included a method for finding an automatic segmentation of the anomalies, as this method is part of the paper. But it is unused in the current code.

Is there a good way to include it as part of the default outputs? I saw that all the segmentations are actually obtained by maximizing some metric in testing sets, so I'm not sure if it's a good idea to try to add this method there... I guess not, but I can't think of any other place.

Or maybe it is just fine as it is... The code is there in case anyone wants to use it...

Any thoughts are appreciated. Thanks!

Maybe you can create an issue for this, and we can refactor this so that users might be able to switch between both. I am not sure what's the priority for this, but it will remain visible until when we finally get around to doing this.

[mtailanian]

@mtailanian the PR looks good to me. Do you have access to the logs here: https://github.com/openvinotoolkit/anomalib/actions/runs/9483018344/job/26129442354?pr=2104? If so, can you fix the pre-merge checks. Ideally, you should be able to run these locally as well. Just do pip install pre_commit and then run pre-commit run --all-files

Hi @ashwinvaidya17 , thank you for your response!
I had already run the pre-commit. However, I'm a bit confused. pre-commit suggest to change the typing, for example list -> List, or tuple -> Tuple. I changed it, but the pre-commit itself would revert it to list, tuple, etc...

How should I do? I'm kind of stuck...

Thanks again!

[samet-akcay]

@mtailanian thanks for your patience. It took me quite some time to review this PR. I have two comments below

[mtailanian]

Hi @samet-akcay , thanks for your feedback. I just addressed both comments!
Let me know if there is something else...
Best,

[samet-akcay]

Hi @samet-akcay , thanks for your feedback. I just addressed both comments! Let me know if there is something else... Best,

Thanks @mtailanian

[mtailanian]

@samet-akcay I see I have again the same pre-commit errors. Please see this comment. I don't really know how to solve it...

[samet-akcay]

It's alright, I'll merge it to a branch, from which I'll try to fix it

[mtailanian]

Thanks so much @samet-akcay !
:)