Table 2.

Our algorithm's performance compared with the state-of-the-art methods for the ETHMS and CAVIAR datasets.

Dataset	Algorithm	False positives per image
ETHMS	Zhang's method 2	0.97
	KSP	0.86
	SSP	0.89
	SPFA	0.77
CAVIAR	Zhang's method 2	0.105
	KSP	0.057
	SSP	0.636
	SPFA	0.051