can you tell me what's the meaning of these code in test.py?

[wzxyks]

test.py： classes_num_filtered = np.argmax(filter_mat_probs, axis=3)[filter_mat_boxes[0], filter_mat_boxes[1], filter_mat_boxes[2]] why not: classes_num_filtered = np.argmax(probs, axis=3)[filter_mat_boxes[0], filter_mat_boxes[1], filter_mat_boxes[2]] ？？？ I can't understand, Please help, thank you.

[Polaris-SDU]

When we do prediction,we should do: step1: Setting some lower probability prediction ( class-specific confidence score for each bounding box<0.2) to zero step2: Sorting from big to small according to confidence score and then calculate IOU

so by the step1: classes_num_filtered = np.argmax(filter_mat_probs, axis=3)[filter_mat_boxes[0], filter_mat_boxes[1], filter_mat_boxes[2]] (I am not a English speaker,it maybe not clear)