EURASIP Journal on Image and Video Processing

Table 2 Evaluation of cost function used to compute thresholds β _j when 60% of weak classifiers are missing

From: Handling missing weak classifiers in boosted cascade: application to multiview and occluded face detection

	k	Cost function	FP	TP	$\bar{n_{level}}$
			×10 ^-3
		TP_cost	8.4	0.95	1.64
P_boost	-	FP_cost	0.058	0.06	1.48
		FP_TP_cost	0.17	0.29	1.49
		TP_cost	8.25	0.96	1.32
P_knn	3	FP_cost	0.15	0.56	1.26
		FP_TP_cost	0.28	0.58	1.32
		TP_cost	11.9	0.97	1.67
P_comb	3	FP_cost	0.005	0.11	1.48
		FP_TP_cost	0.16	0.49	1.5

Three evaluation terms are exposed: the false positive rate, the true positive rate and the mean number of evaluated levels per negative example noted $\bar{n_{level}}$ .

Back to article page