Weighted principal component analysis
AICI'11 Proceedings of the Third international conference on Artificial intelligence and computational intelligence - Volume Part III
Hi-index | 0.00 |
Because of the low recognition rate by affinity propagation (AP) we propose an algorithm to make face recognition based on two dimension double PCA (2DDPCA) and affinity propagation in this paper. First we use two dimension principle component analysis (2DPCA) to compress the face images in horizontal direction, then we compress the features in vertical direction using 2DPCA again. At last we use affinity propagation classifying faces on the basis of face database. Compared to the original affinity propagation, the superiority of this method is that the speed of classification of large sample is faster and the recognition rate stabilized at a high level. The test results of ORL Face Database and Yale Face Database prove the effectiveness of this method.