When we are training Tensorflow Object Detection API, in Tensorboard there are 3 metrics for recall:
- 'DetectionBoxes_Recall/AR@1': average recall with 1 detection.
- 'DetectionBoxes_Recall/AR@10': average recall with 10 detections.
- 'DetectionBoxes_Recall/AR@100': average recall with 100 detections.
How are those calculated? What 'with n detections' exactly mean?
