An Appearance Invariant Gait Recognition Technique Using Dynamic Gait Features

Table 4

Gait recognition accuracy achieved by our work and existing work.

Research work	CASIA-B	TUM-IITKGP	OUISIR-B
TTGS + MCCNN [26]	99%	—	—
3D Gait model + partial Similarity [22]	99%	99%	—
	96%	80%	—
	95%	65%
	Average = 96.6%
3D Gait + Sparse reconstruction [23]	96%	—	—
GEI + PCA + WRSL [12]	89%	—	—
GEI + DRL + CNN [17]	92.6%	—	—
GEI + MSCNN [38]	90.43%	—	—
Effective joints + LSTM + CNN [27]	96%	—	—
	79%
	61%
	Average = 79.6%
Pose + LSTM + CNN [39]	97.58%	—	—
	70.16%
	56.45%
	Average = 74.7%
Optical flow, PCA, LDA [40]	98%
	90%
	64%
	Average = 84%
Our work (DGF, CCS, SVM)	98%	97.1%	100%