lazyslide.metrics.segmentation.recall

recall(stats)

Compute recall (sensitivity) from segmentation statistics.

Recall measures the proportion of actual positive instances that were correctly identified. It answers the question: “Of all the instances that are actually positive, how many did we correctly predict as positive?”

Parameters:

statsSegmentationStats

Segmentation statistics containing true positives (tp) and false negatives (fn).

Returns:

float

Recall score in the range [0, 1], where 1 is perfect recall. A small epsilon (1e-8) is added to the denominator to avoid division by zero.

Notes

Recall = TP / (TP + FN) Also known as sensitivity or true positive rate (TPR).