dgs.models.similarity.pose_similarity.IntersectionOverUnion.get_target

IntersectionOverUnion.get_target(ds: State) → torchvision.tv_tensors.BoundingBoxes

Given a State obtain the ground-truth bounding-boxes as torchvision.tv_tensors.BoundingBoxes object of size [T x 4].

Notes

The function box_iou() expects that the bounding boxes are in the ‘XYXY’ format.