A Gabor-Block-Based Kernel Discriminative Common Vector Approach Using Cosine Kernels for Human Face Recognition

Table 1

Average recognition results using FRAV2D database.

	Recognition accuracy (%)
Method	No. of training samples		Average recognition rates (%)
	3	4
GWT	85.5	89.5	87.5
KDCV	79.5	82	80.75
GWT-LDA	88.3	90.33	89.33
GWT-KDCV (RBF)	87	90	88.5
GWT-KDCV (Cosine)	88	90	89
GWT-KDCV (RBF, low-energized)	90	92.5	91.25
GWT-KDCV (Cosine, low-energized)	96.125	97.75	96.9