Scientific Reports

Table 2 Quantitative comparison on SOD³⁹, DUT-OMRON⁴⁰ and PASCAL-S⁴¹ datasets.

From: Salient object detection with non-local feature enhancement and edge reconstruction

Method	SOD					DUT-OMRON					PASCAL-S
Method	\(S_{m}~\uparrow\)	\(F^{\omega }_{\beta }~\uparrow\)	\(MAE~\downarrow\)	\(F^{max}_{\beta }~\uparrow\)	\(E^{max}_{m}~\uparrow\)	\(S_{m}~\uparrow\)	\(F^{\omega }_{\beta }~\uparrow\)	\(MAE~\downarrow\)	\(F^{max}_{\beta }~\uparrow\)	\(E^{max}_{m}~\uparrow\)	\(S_{m}~\uparrow\)	\(F^{\omega }_{\beta }~\uparrow\)	\(MAE~\downarrow\)	\(F^{max}_{\beta }~\uparrow\)	\(E^{max}_{m}~\uparrow\)
PiCANet	0.793	0.723	0.109	0.858	0.866	0.832	0.695	0.065	0.803	0.876	0.854	0.780	0.087	0.881	0.901
BASNet	0.772	0.728	0.114	0.851	0.832	0.836	0.751	0.056	0.805	0.871	0.838	0.793	0.076	0.854	0.886
SCRN	0.792	0.734	0.104	0.867	0.863	0.836	0.720	0.056	0.812	0.875	0.869	0.807	0.064	0.882	0.910
LDF	0.800	0.765	0.093	0.873	0.866	0.839	0.752	0.051	0.820	0.869	0.863	0.822	0.060	0.874	0.908
U2Net	0.786	0.748	0.108	0.861	0.857	0.847	0.757	0.054	0.823	0.880	0.844	0.797	0.074	0.859	0.883
ITSD	0.809	0.777	0.095	0.880	0.874	0.840	0.750	0.061	0.824	0.880	0.859	0.812	0.071	0.871	0.908
CTDNet	–	–	–	–	–	0.844	0.762	0.052	0.826	0.881	0.863	0.822	0.061	0.878	0.906
VST	0.854	0.778	0.065	0.866	0.902	0.850	0.755	0.058	0.800	0.888	0.873	0.816	0.067	0.850	0.900
RCSB	0.750	0.730	1.536	0.846	0.813	0.835	0.752	0.045	0.809	0.866	0.860	0.816	0.058	0.876	0.906
EDN	0.798	0.767	0.252	0.868	0.864	0.838	0.746	0.057	0.805	0.871	0.860	0.815	0.066	0.875	0.903
ICON	0.814	0.784	0.089	0.872	0.866	0.833	0.743	0.065	0.817	0.879	0.861	0.820	0.064	0.878	0.911
OLER	–	–	–	–	–	0.845	0.775	0.054	0.826	0.891	0.858	0.827	0.063	0.877	0.906
MENet	0.809	0.777	0.087	0.878	0.864	0.850	0.771	0.045	0.834	0.879	0.872	0.838	0.054	0.890	0.915
ELSANet	–	–	–	–	–	0.846	0.774	0.050	0.794	0.885	0.864	0.836	0.060	0.862	0.910
EMSNet	0.798	0.819	1.962	0.863	0.852	0.838	0.759	0.048	0.807	0.870	0.868	0.848	0.054	0.892	0.915
DC-Net	0.797	0.761	0.100	0.862	0.855	0.849	0.772	0.053	0.827	0.883	0.857	0.818	0.067	0.878	0.899
CANet	0.803	0.777	0.244	0.874	0.869	0.847	0.828	0.047	0.827	0.886	0.859	0.888	0.061	0.876	0.910
Ours	0.824	0.801	0.082	0.886	0.881	0.860	0.789	0.051	0.839	0.899	0.874	0.837	0.057	0.882	0.919

For \(\uparrow\) and \(\downarrow\), higher and lower scores indicate better results, respectively. \(E_m^{max}\) denotes max E-measure, \(F_\beta ^{max}\) denotes max F-measure. The best and second-best results are shown in bold and italics, respectively. The symbol ’–’ indicates the results of the model are unavailable.

Back to article page

Search

Advanced search

Quick links