GdoongMathew

Results 51 comments of GdoongMathew

@AL3708 and @Dipet I've implemented some of the bboxes transforms with numpy in [my repo](https://github.com/GdoongMathew/albumentations). But there's a little problem in this, unless you have more then 20 bounding boxes...

In albumentations, usually an image or mask is defined in channel-last style, thus you'll have to transpose it like ```python image = image.transpose(1, 2, 0) ```

@ynhuhu did you use any thing like transpose or ToTorchV2 layers in your Compose? or probably you can provide that piece of code? Thanks~

I've added a few fixes. 1. converting bboxes / keypoints to internal type when testing transforming function. 2. add targets back in the unittests. 3. some requested changes. point 1...

@Dipet Hi, changes have been made as requested.

Hi @Dipet, since I haven't got further comments on [this change request](https://github.com/albumentations-team/albumentations/pull/1396#discussion_r1103777429), I think currently it's best to just add the correct type hint instead of combining `combine_bboxes_from_albumentations`, `convert_to_original_type` &...

Hi @Dipet I think there was a mistake when merging master into this branch, accidentally renamed `convert_bboxes_to_albumentations` to `a` in bbox_utils.py?

Hi @Dipet I think the version of scikit-image is the reason why the unittest failed. Should I fix its version to `scikit-image>=0.16.1,

@Dipet Hi, is there anything that still needs to be changed or that I can help with?