Table 9 Segmentation Performance Comparison (ISIC 2018 Dataset).

From: HyperFusionNet combines vision transformer for early melanoma detection and precise lesion segmentation

Method

Dice coefficient

Jaccard index

Time (s)

U-Net 39

0.850

0.790

0.96

Attention U-Net 38

0.875

0.815

1.13

DeepLabV3 + 42

0.870

0.810

1.25

TransUNet 43

0.890

0.830

1.68

Swin-UNet 44

0.905

0.845

1.92

HyperFusion-Net (Proposed)

0.932

0.881

1.15

Back to article page