Table 2.

Kvasir-SEG evaluation results of all methods.

Method	Params	Dice	IoU	Recall	Precision
A-DenseUNet	11.0 M	0.9085	0.8615	0.9448	0.9766
UNet++ [11]	9.0 M	0.8021	0.7215	0.7914	0.9321
ResUNet [40]	8.5 M	0.7864	0.5421	0.7861	0.8912
Wide U-Net [11]	9.1 M	0.7645	0.7112	0.7684	0.9231
U-Net [10]	7.8 M	0.7062	0.5628	0.7768	0.9022
PraNet [52]	-	0.898	0.84	-	-
ResUNet++ [6]	-	0.8133	0.7927	0.7064	0.8774