Multispectral Local Binary Pattern Histogram for Component-based Color Face Verification

A novel discriminative face representation derived by the Linear Discriminant Analysis (LDA) of multispectral Local binary pattern histograms is proposed for color face recognition. The color face image is first photometrically normalized and partitioned into several non-overlapping regions. In each region, multispectral local binary pattern histograms are extracted and concatenated into a regional feature. The feature is then projected into a LDA space to be used as a regional discriminative facial descriptor. The overall similarity score is obtained by fusing the similarity scores of the regional descriptors. The method is implemented and tested in face verification on the XM2VTS and FRGC 2.0 databases with very promising results.