Can any tlt object detection models support rotated bounding box annotations and can it produce detections with same rotation. I have a data where I should use rotated bounding boxes for better detection.
In detectnet_v2, see DetectNet_v2 — Transfer Learning Toolkit 3.0 documentation, there is configurable parameter in Augmentation Module.
rotate_rad_max : The angle of rotation to be applied to the images and the training labels. The range is defined between [-rotate_rad_max, rotate_rad_max].
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.