Table 4 Comparison with baseline models.

From: An improved lightweight tongue segmentation model with self-attention parallel network and progressive upsampling

 

IoU

Acc

Dice

Precision

Recall

Vit

93.23

96.73

96.50

96.27

96.73

HRNet

94.12

95.99

96.97

97.97

95.99

Swin Transformer

95.80

97.56

97.86

98.15

97.56

Beit

91.77

92.79

95.71

93.72

97.79

PAPU_TonSeg

96.82

98.21

98.38

98.55

98.21

Back to article page