Research Article

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

Table 6

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

MethodsC1C2C3C4

LDA21.45 ± 1.2220.76 ± 1.5721.15 ± 1.0620.77 ± 1.39
PCA43.49 ± 1.4644.16 ± 0.6443.98 ± 1.1045.03 ± 1.42
LatLRR33.71 ± 1.1334.92 ± 0.8830.73 ± 1.2833.55 ± 1.36
LPP35.13 ± 1.0431.26 ± 0.9732.72 ± 1.6034.16 ± 1.31
SRRS75.64 ± 1.0266.72 ± 0.8470.56 ± 1.1781.14 ± 0.45
RCVL73.20 ± 0.9464.96 ± 0.4866.03 ± 0.5679.02 ± 1.26
Ours76.25 ± 0.5768.36 ± 0.2471.69 ± 1.1082.64 ± 1.16