FACE RECOGNITION USING TRANSFORM FEATURES AND NEURAL NETWORKS

In this paper we considered face recognition using two Radial Basis Fu nction Network (RBFN) architectures and compared performance with the nearest neighbor algorithm. Performance was also evaluated for feature vectors extracted from face images by using principal component analy sis as well as wavelet transform. Raw recognition rates as well as rat es with confidence measures were considered. in the RBFN1 architecture , one network was used to discriminate among the classes, while the RB FN2 architecture used one network per class. From the point of view of computations RBFN2 was more efficient than RBFN1 with PCA feature vec tors, but its recognition performance was slightly worse than RBFN1. O ther experiments showed that RBFN1 was largely superior to the NNA whe n the amount of computations in both methods was similar. With the use of wavelet features, performance dropped 5-10% in relation to feature s extracted by PCA. However, in a given implementation, this must be w eighed in conjunction with the advantages of using wavelet features, n amely, no storage is required for eigenvectors, and they are simpler t o compute. (C) 1997 Pattern Recognition Society. Published by Elsevier Science Ltd.