Skip to main content

Table 3 Mean average precisions obtained by Yolov5 with different confidence thresholds

From: Application of deep learning in wound size measurement using fingernail as the reference

Confidence threshold

mAP@0.501

mAP@0.752

mAP@0.5:0.953

0.1

0.819

0.797

0.704

0.2

0.819

0.805

0.723

0.3

0.822

0.806

0.725

0.4

0.818

0.804

0.723

0.5

0.811

0.798

0.715

0.6

0.804

0.799

0.715

0.7

0.799

0.795

0.713

0.8

0.782

0.778

0.704

0.9

0.494

0.494

0.478

  1. (1) mAP value at IoU = 0.5, (2) mAP value at IoU = 0.75, (3) average mAP over different IoU thresholds (0.5 to 0.95 in steps of 0.5)