Skip to main content

Table 1 The comparisons of the rock-art segmentation performance of BEGL-UNet with various loss functions on the test set

From: BEGL: boundary enhancement with Gaussian Loss for rock-art image segmentation

Methods

Accuracy

Precision

Recall

F1-score

MIoU

DSC

BE-UNet

0.695

0.450

0.964

0.613

0.514

0.613

FL-UNet

0.757

0.899

0.056

0.104

0.404

0.104

DL-UNet

0.829

0.591

0.992

0.739

0.679

0.739

BCE-UNet

0.907

0.804

0.832

0.804

0.778

0.804

BEGL-UNet

0.935

0.863

0.873

0.865

0.840

0.865

  1. BEGL-UNet achieves the best results on Accuracy (0.935), F1 (0.865), MIoU (0.840) and DSC (0.865), only the worse results on precision and recall which are competitive with the best results