Table 7 Comprehensive performance comparison across five benchmark datasets using F-measure (\(\hbox {F}^{w}_{\beta }\)), MAE, S-measure (S\(\alpha\)), and mean E-measure (mE\(\xi\)) metrics.

Models	Endoscene				ClinicDB				ColonDB				ETIS				Kvasir-SEG
	\(\hbox {F}^{w}_{\beta }\)	MAE	S\(\alpha\)	mE\(\xi\)	\(\hbox {F}^{w}_{\beta }\)	MAE	S\(\alpha\)	mE\(\xi\)	\(\hbox {F}^{w}_{\beta }\)	MAE	S\(\alpha\)	mE\(\xi\)	\(\hbox {F}^{w}_{\beta }\)	MAE	S\(\alpha\)	mE\(\xi\)	\(\hbox {F}^{w}_{\beta }\)	MAE	S\(\alpha\)	mE\(\xi\)
UNet	0.684	0.022	0.843	0.848	0.811	0.019	0.889	0.913	0.491	0.059	0.710	0.692	0.366	0.036	0.684	0.643	0.794	0.055	0.858	0.881
UNet++	0.687	0.018	0.839	0.834	0.785	0.022	0.873	0.891	0.467	0.061	0.692	0.680	0.390	0.035	0.683	0.629	0.808	0.048	0.862	0.886
PraNet	0.843	0.010	0.925	0.950	0.896	0.009	0.936	0.963	0.699	0.043	0.820	0.847	0.600	0.031	0.794	0.808	0.885	0.030	0.915	0.944
ACSNet	0.825	0.013	0.923	0.939	0.873	0.011	0.927	0.947	0.697	0.039	0.829	0.839	0.530	0.059	0.754	0.737	0.882	0.032	0.920	0.941
Polyp-PVT	0.884	0.007	0.935	0.973	0.936	0.006	0.949	0.985	0.795	0.031	0.865	0.913	0.750	0.013	0.871	0.906	0.911	0.023	0.925	0.956
BDG-Net	0.876	0.006	0.937	0.967	0.905	0.007	0.938	0.970	0.714	0.015	0.866	0.895	0.776	0.031	0.866	0.894	0.896	0.028	0.918	0.952
SSFormer-L	0.875	0.007	0.939	0.969	0.906	0.008	0.934	0.963	0.790	0.031	0.866	0.901	0.761	0.015	0.881	0.905	0.911	0.023	0.923	0.957
PVT-CASCADE	0.882	0.008	0.934	0.965	0.923	0.013	0.939	0.969	0.798	0.029	0.864	0.910	0.775	0.016	0.886	0.906	0.918	0.020	0.928	0.964
MEGANet-ResNet	0.863	0.009	0.924	0.956	0.931	0.008	0.950	0.977	0.766	0.038	0.845	0.897	0.753	0.015	0.866	0.912	0.904	0.026	0.916	0.952
PVT-EMCAD-B2	0.869	0.007	0.921	0.965	0.927	0.010	0.943	0.973	0.804	0.028	0.869	0.919	0.759	0.016	0.877	0.902	0.920	0.021	0.929	0.966
BiCoMA (ours)	0.892	0.006	0.946	0.979	0.928	0.005	0.949	0.985	0.829	0.019	0.878	0.923	0.788	0.018	0.889	0.922	0.927	0.018	0.936	0.958

Quick links

Search