Research and implementation of of multi-modal face recognition algorithm
Jihua Ye, Xia Guomiao, Dan Hu · 2013
Traditional face recognition algorithm based on two-dimensional face recognition; Due to the influence of illumination, changes in the expression and Real human face image extraction from different angles, the recognition of such algorithms rate is considerably limited. To solve these problems, a multi-modal face recognition algorithm is presented in this paper. After the Feature extraction using 2DPCA and the reduction of the dimension from multiple two-dimensional face images (a positive face and two side faces) of one person, and to identify it by a new feature Matrix which was reformed by part of the feature vector of each two-dimensional face images. The experimental result of CAS-PEAL Face Database shows that this method receives a higher recognition rate than the method which use a single positive face under the same conditions.