search

rafaelpadilla / review_object_detection_metrics

Object Detection Metrics. 14 object detection metrics: mean Average Precision (mAP), Average Recall (AR), Spatio-Temporal Tube Average Precision (STT-AP). This project supports different bounding box formats as in COCO, PASCAL, Imagenet, etc.
Other
1.09k stars 215 forks source link

Is it possible to have a FP and FN at the same time? #60

Closed dariogonle closed 3 years ago

dariogonle commented 3 years ago

The green box is the ground truth and the red the predicted box. If I set the iou threshold to 0.5 I would consider this prediction as a bad prediction, so it would be a FP. But as the ground truth is not detected would it count as well as a FN?

I mean does this count as FP, or as a FP and FN?

image

Thank you.

rafaelpadilla commented 3 years ago

@dariogonle As your detection (positive) missed the object, it will be counted as FP and, as your groundtruth is not detected by any other positive, it will be counted as FN.

Observations: Note that to calculate the Precision and Recall, the FN appears in the denominator of the Recall, as Recall = TP/(TP+FN), where TP+FN=all ground truths. So, in fact, you dont have to compute FN, you just have to consider 'all ground truths'.

dariogonle commented 3 years ago

Thank you.