Table 5 EER (%) for thinned view angle pairs for a cooperative setting

From: Multi-view large population gait dataset and its performance evaluation for cross-view gait recognition

(a)VTM						(b)wQVTM
	Probe						Probe
Gallery	0	30	60	90	Mean	Gallery	0	30	60	90	Mean
0	6.9	34.2	40.0	40.7	30.5	0	5.1	30.1	37.8	38.9	28.0
30	32.3	6.0	27.3	31.9	24.4	30	30.8	5.1	22.0	31.3	22.3
60	35.8	22.9	7.1	21.5	21.8	60	34.8	19.1	6.0	16.9	19.2
90	36.2	29.0	22.7	5.9	23.4	90	34.5	25.8	17.7	4.5	20.6
Mean	27.8	23.0	24.3	25.0	25.0	Mean	26.3	20.1	20.9	22.9	22.5
(c)LDA						(d)1in-GEINet
	Probe						Probe
Gallery	0	30	60	90	Mean	Gallery	0	30	60	90	Mean
0	7.1	27.3	38.3	40.0	28.2	0	3.4	7.4	16.8	17.0	11.2
30	27.4	5.5	22.1	32.8	21.9	30	7.7	1.9	5.4	8.6	5.9
60	38.8	22.4	6.6	18.6	21.6	60	16.8	5.3	2.4	4.6	7.3
90	40.2	32.8	18.5	5.5	24.2	90	17.4	8.6	4.6	1.7	8.1
Mean	28.4	22.0	21.4	24.2	24.0	Mean	11.3	5.8	7.3	8.0	8.1
(e) Siamese						(f) MT
	Probe						Probe
Gallery	0	30	60	90	Mean	Gallery	0	30	60	90	Mean
0	2.5	3.9	6.8	6.0	4.8	0	2.3	6.3	11.2	10.7	7.6
30	4.0	1.4	2.9	2.8	2.8	30	6.3	1.4	3.4	4.8	4.0
60	7.0	2.7	1.4	2.5	3.4	60	11.4	3.5	1.3	3.1	4.8
90	6.2	2.7	2.6	1.1	3.1	90	11.5	5.1	3.0	1.1	5.2
Mean	4.9	2.7	3.4	3.1	3.5	Mean	7.9	4.1	4.7	4.9	5.4
(g) LB						(h) DM
	Probe						Probe
Gallery	0	30	60	90	Mean	Gallery	0	30	60	90	Mean
0	2.3	6.4	12.1	11.7	8.1	0	6.9	26.8	43.3	46.0	30.8
30	6.8	1.5	4.0	5.2	4.4	30	27.3	6.0	28.4	39.7	25.4
60	12.8	4.2	1.5	3.2	5.4	60	43.2	29.2	7.1	19.9	24.9
90	12.6	5.7	3.2	1.0	5.6	90	46.5	39.6	19.8	5.9	27.9
Mean	8.6	4.5	5.2	5.3	5.9	Mean	31.0	25.4	24.6	27.9	27.2

Back to article page