I'm labelling objects for object detection with YOLOv5 and tried using makesense.ai, labelImg,... for that. It works fine with horizontal/vertical objects, but I do have some vertical ones in my dataset. Is there a way for diagonal Bounding Boxes? Or is there any other way for labelling diagonal objects for improving performance of my model? Thanks!
Yes, but your bounding box will be squared or rectangular. For instance, the image below contains a small crack with a diagonal shape (red box). Unfortunately, YOLOv5 does not support other kinds of bounding boxes yet.