Recognizing Faces Using Kernel Eigenfaces

In face recognition, Principal Component Analysis (PCA) is often used to extract a low dimensional face representation based on the eigenvector of the face image autocorrelation matrix. Kernel Principal Component Analysis (Kernel PCA) has recently been proposed as a non-linear extension of PCA. While PCA is able to discover and represent linearly embedded manifolds, Kernel PCA can extract low dimensional non-linearly embedded manifolds from data, thus providing a more suitable recognition by a classifier. We provide experimental evidence which show that Kernel PCA performs better than PCA on the ATT Face Dataset when both are used with a lienar Support Vecter Machine Classifier. Philippine Computing Journal 1(1) 2006 27-30