An Appearance Invariant Gait Recognition Technique Using Dynamic Gait Features

Table 3

Standard deviation and relative standard deviation score of CCS for gait verification.

Dataset	Pair (Use Case 1, Use Case 2)	Mean	Variance	SD	RSD = SD/mean %
CASIA-B	Pair 1 (normal, bag)	0.4	0.015	0.12	0.12/0.4 = 0.3 = 30%
	Pair 2 (normal, long coat)	0.4	0.014	0.2	0.2/0.4 = 0.5 = 50%
	Pair 3 (bag, long coat)	0.6	0.04	0.2	0.2/0.6 = 0.33 = 33%
OUISIR-B	Pair 1 (normal, loose)	0.2	0.02	0.141	0.141/0.2 = 0.7 = 70%
	Pair 2 (normal, long coat)	0.19	0.02	0.144	0.144/0.19 = 0.7 = 70%
	Pair 3 (loose, long coat)	0.16	0.02	0.14	0.14/0.16 = 0.8 = 80%
TUM-IITKGP	Pair 1 (normal, back pack)	0.29	0.03	0.18	0.18/0.29 = 0.6 = 60%
	Pair 2 (normal, gown)	0.27	0.02	0.14	0.14/0.27 = 0.5 50%
	Pair 3 (back pack, gown)	0.28	0.02	0.15	0.15/0.28 = 0.5 = 50%
SACV	Pair 1 (fitted, fitted with bag)	0.42	0.05	0.23	0.23/0.42 = 0.54 = 54%
	Pair 2 (fitted, knee down)	0.44	0.048	0.22	0.22/0.44 = 0.5 = 50%
	Pair 3 (knee down, knee down with bag)	0.52	0.056	0.23	0.23/0.52 = 0.44 = 44%
	Pair 4 (fitted with bag, knee down with bag)	0.47	0.038	0.19	0.19/0.47 = 0.4 = 40%