Hierarchical Multimodal Adaptive Fusion (HMAF) Network for Prediction of RGB-D Saliency

Table 3

The evaluation results of various saliency prediction models.

Datasets	Criteria	Fang	DeepFix	DVA	SAM	EML	SMI	Proposed
NUS	CC	0.333	0.4322	0.4549	0.5013	0.4857	0.5368	0.5533
	KLDiv	1.560	1.8138	2.4349	2.9059	2.2353	1.2501	1.0794
	AUC	0.795	0.7699	0.7236	0.7461	0.8149	0.8648	0.8422
	NSS	1.209	1.6608	1.7962	2.1259	1.9419	2.1843	2.3069
NCTU	CC	0.542	0.7974	0.6834	0.7115	0.7556	0.8376	0.8472
	KLDiv	0.674	1.3083	1.1045	1.8482	0.3886	0.4826	0.3228
	AUC	0.806	0.8650	0.8035	0.7250	0.8818	0.9053	0.9130
	NSS	1.264	1.8575	1.5546	1.5386	2.0666	2.3121	2.3493