Table 6.

Performance (Dice) of different approaches in 10 datasets for skin detection. The bold represents the best performance.

	DA	PRAT	MCG	UC	CMQ	SFA	HGR	SCH	VMD	ECU	VT	AVG
H_S	DA1	0.903	0.880	0.903	0.838	0.947	0.964	0.793	0.744	0.941	0.810	0.872
H_S	DA2	0.911	0.884	0.903	0.844	0.950	0.968	0.776	0.683	0.943	0.835	0.870
H_A	DA1	0.913	0.880	0.900	0.809	0.951	0.967	0.792	0.717	0.945	0.799	0.867
H_A	DA2	0.909	0.886	0.893	0.848	0.951	0.968	0.775	0.707	0.944	0.832	0.871
FH(2)	DA1/DA2	0.920	0.892	0.913	0.859	0.953	0.971	0.793	0.746	0.951	0.839	0.884
FH(4)	DA1/DA2	0.920	0.892	0.916	0.862	0.954	0.971	0.795	0.765	0.951	0.831	0.886
PVT	DA1	0.920	0.888	0.925	0.851	0.951	0.966	0.792	0.709	0.951	0.828	0.878
PVT	DA2	0.923	0.892	0.908	0.863	0.951	0.968	0.776	0.709	0.952	0.848	0.879
PVT(2)	DA1/DA2	0.925	0.892	0.925	0.863	0.952	0.970	0.781	0.719	0.954	0.850	0.883
HSN	DA1	0.927	0.893	0.920	0.851	0.953	0.966	0.777	0.704	0.951	0.800	0.874
HSN	DA2	0.924	0.896	0.889	0.860	0.953	0.969	0.781	0.690	0.953	0.855	0.877
HSN(2)	DA1/DA2	0.928	0.897	0.915	0.860	0.955	0.970	0.775	0.671	0.953	0.860	0.879
FH(2) + 2 × PVT(2)	DA1/DA2	0.927	0.894	0.932	0.868	0.954	0.971	0.797	0.767	0.955	0.853	0.893
FH(4) + 4 × PVT(2)	DA1/DA2	0.926	0.894	0.933	0.869	0.954	0.971	0.798	0.768	0.955	0.847	0.892
ElossMix2(10)	DA1/DA2	0.924	0.893	0.929	0.850	0.956	0.970	0.789	0.739	0.952	0.829	0.883
AllM	DA1/DA2	0.929	0.895	0.939	0.868	0.956	0.972	0.800	0.770	0.956	0.846	0.893
AllM_H	DA1/DA2	0.931	0.897	0.941	0.869	0.956	0.972	0.799	0.773	0.957	0.854	0.895