Connect Oriented Bounding Box to Metrics

[LinasKo]

Connect Oriented Bounding Box to Metrics

Tip

Hacktoberfest is calling! Whether it's your first PR or your 50th, you’re helping shape the future of open source. Help us build the most reliable and user-friendly computer vision library out there! 🌱

---

Several new features were recently added to supervision:

• Mean Average Precision (mAP)
• F1 Score
• IoU calculation for Oriented Bounding Boxes

Intersection Over Union (IoU) is the starting point when computing these metrics. It determines which detections are considered true positives. However, take a look! The Oriented Box IoU is not supported yet! Help us add support by using oriented_box_iou_batch.

Helpful links:

• Contribution guide
• Metrics:
  • mAP metric: docs, code
  • F1 Score: docs, code
• Oriented box IoU calculation function: docs, code
• Supervision Cheatsheet
• Colab Starter Template
• Prior metrics test Colab

[LinasKo]

Hi @patel-zeel 👋

I'm moving the IoU integration into metrics here. I believe we spoke about it briefly. Have you done substantial work on it already? If not, I will open it to the community.

[LinasKo]

Contribution guidelines

If you would like to make a contribution, please check that no one else is assigned already. Then leave a comment such as "Hi, I would like to work on this issue". We're happy to answer any questions about the task even if you choose not to contribute.

Please share a Google Colab with minimal code to test the new feature. We know it's additional work, but it will speed up the review process. You may use the Starter Template. The reviewer must test each change. Setting up a local environment to do this is time-consuming. Please ensure that Google Colab can be accessed without any issues (make it public). Thank you! 🙏

[Prasadayus]

Hi ,I would like to work on this issue. Please @LinasKo assign me this issue

[patel-zeel]

Hi @LinasKo, feel free to open it for the community. I'd be happy to help other contributors if they face any issues while implementing :)

[onuralpszr]

Hi ,I would like to work on this issue. Please @LinasKo assign me this issue

@Prasadayus hello task is yours, good luck have fun :)

[abhinav7289A]

Hi @LinasKo I am new to open source contribution. Please assign me this issue as I want to excel in this field

[LinasKo]

Hi @abhinav7289A,

This issue is already taken. I'll add many new ones next week - keep an eye on our issue board! 😉

[PrakharJain1509]

If an issue is already taken and I propose a better solution, then will you consider it or not? @LinasKo

[LinasKo]

Hi @PrakharJain1509 👋
No. However, after the initial solution is refined and merged, I'm happy to review PRs that meaningfully improve upon it, even if the issue is closed. This would be eligible for Hacktoberfest as well.

[LinasKo]

Hi @Prasadayus 👋

How's it going? Did you find any time to look at oriented boxes and metrics? 😉

[Prasadayus]

@LinasKo Sorry for the late reply, I was very busy this week. I cloned the project, reviewed the oriented box and will make the changes by the end of this week

[LinasKo]

Thank you very much!

[PrakharJain1509]

@LinasKo I have opened a PR for this issue. Please review it and let me know if any changes are required.

[LinasKo]

Thank you very much @PrakharJain1509 ! I'll have a look and review it tomorrow

[LinasKo]

I've just realized that @Prasadayus is the original assigned contributor. Your PR will have priority over that of @PrakharJain1509.

@Prasadayus, how goes it? Do you think you'll have time to complete this PR? 😉

[LinasKo]

Adding @PrakharJain1509 as an assignee.
Of the two, the first successfully merged contribution will be chosen.

[LinasKo]

Accepting #1593 as the solution for adding OBB support for the mAP metric. It needs a lot of cleaning-up, but I like the effort put in.

This ticket also involves the F1 metric. @Prasadayus, will you have the time to look into it?

[rahulsamant37]

@LinasKo Sir, Can you review my changes?

[LinasKo]

@rahulsamant37, I reviewed your changes in the PR and decided to close it in favor of #1593, as it aligns better with our current direction. Thanks for your effort and looking forward to your future contributions!

[LinasKo]

Completed via #1593. Thank you for your contributions, everyone! I hope you found the experience useful and fun!