Table 4 Experimental results in comparison with other algorithms.

Method	IOU	Dice	ACC	RE	PR
Kvasir-SEG
U-Net	0.749	0.821	0.923	0.817	0.825
U-Net + +	0.752	0.825	0.927	0.824	0.831
DeepLabV3	0.801	0.876	0.956	0.923	0.841
PraNet	0.835	0.896	0.973	0.915	0.885
TransFuse	0.784	0.873	0.957	0.898	0.862
TransUnet	0.796	0.884	0.962	0.905	0.873
Ours	0.823	0.892	0.975	0.917	0.909
CVC-ClinicDB
U-Net	0.727	0.825	0.923	0.827	0.831
U-Net + +	0.734	0.837	0.935	0.841	0.827
DeepLabV3	0.748	0.849	0.968	0.879	0.836
PraNet	0.849	0.899	0.982	0.936	0.896
TransFuse	0.765	0.852	0.978	0.885	0.843
TransUnet	0.837	0.909	0.979	0.895	0.874
Ours	0.851	0.893	0.985	0.907	0.889

Quick links

Search