Performs non-maximum suppression in a batched fashion. Each index value correspond to a category, and NMS will not be applied between elements of different categories.
batched_nms(boxes, scores, idxs, iou_threshold)(Tensor[N, 4]): boxes where NMS will be performed. They are expected to be in \((x_{min}, y_{min}, x_{max}, y_{max})\) format with
\(0 \leq x_{min} < x_{max}\) and
\(0 \leq y_{min} < y_{max}\).
(Tensor[N]): scores for each one of the boxes
(Tensor[N]): indices of the categories for each one of the boxes.
(float): discards all overlapping boxes with IoU > iou_threshold
keep (Tensor): int64 tensor with the indices of the elements that have been kept by NMS, sorted in decreasing order of scores