How to replace bbox in coco-annotations with COCO_val2017_detections_AP_H_56_person.json #2198
Replies: 2 comments
-
whether its right to think so , as for top-down regression validatation, it contains two parts, one is about validating bbox and the other is validating keypoints in bbox . COCO_val2017_detections_AP_H_56_person.json contains more than one bbox for one target, thus the bbox with highest score would be used for validation and the internal keypoints would be used for validation following. |
Beta Was this translation helpful? Give feedback.
-
If i'm not wrong, and only the bbox with the highest score in fasterRCNN detected json would used for validation, the results may be light different. it seems like two perspectives of evaluation, i wonder which one is more recommended ? |
Beta Was this translation helpful? Give feedback.
-
Hi ! I have been using mmpose 1.X recently, and there is a question i wanna search for help.
When I was using topdown_regression algorithm with coco datasets, what if i set 'bbox_file=None', and replace the bbox in coco-annotaions with the bbox for the highest score in COCO_val2017_detections_AP_H_56_person.json,.
Would that be equal to setting 'bbox_file=COCO_val2017_detections_AP_H_56_person.json' ?
If not, how can I reach the same result while setting bbox_file to None? Thanks in advance !
@jin-s13
Beta Was this translation helpful? Give feedback.
All reactions