Open JAYCHOU2020 opened 3 years ago
I would also be interested knowing that, with MOT 2d bounding boxes format
"Classification-Aware HOTA" in the paper
I am interested too
I also haven't found any information about it. Any help would be appreciated. The documentation only says that the file should follow this format: frame, id, bb_left, bb_top, bb_width, bb_height, conf, x, y, z
how to evaluate mutil class dataset?