Research Article

Bio-Inspired Structure Representation Based Cross-View Discriminative Subspace Learning via Simultaneous Local and Global Alignment

Table 5

Classification results (%) of all methods on 10% corrupted COIL-20 dataset.

MethodsC1C2C3C4

LDA34.13 ± 1.1135.12 ± 0.8735.27 ± 1.2438.55 ± 1.47
PCA49.44 ± 0.9751.69 ± 0.9448.31 ± 1.4950.81 ± 0.61
LatLRR58.71 ± 1.5857.74 ± 1.4957.66 ± 1.2760.00 ± 1.51
LPP57.63 ± 1.2854.84 ± 1.4258.25 ± 1.3158.62 ± 1.01
SRRS80.72 ± 0.7074.36 ± 0.7975.50 ± 0.4189.06 ± 0.57
RCVL78.20 ± 1.0771.92 ± 1.0273.67 ± 0.8087.13 ± 0.55
Ours82.22 ± 0.9075.14 ± 0.6577.08 ± 0.3190.28 ± 0.93