Performance Comparison of Face Recognition Algorithm Based on Accuracy Rate

Abstract: This paper presents an overview of different face recognition techniques, study and analysis of the face recognition rate, failure rate, training time and recognition time of various face recognition algorithms like PCA, LDA, SVM, ICA and SVD. We analyzed from the results that the performance of SVD in terms of accuracy was better when no. of images increased on ORL database whereasthe performance of PCA was the lowest. Also SVD took more recognition time than other algorithms. Training time of SVM is lowest and PCA, LDA and ICA consume marginal time.

Keywords: Face Recognition, PCA, LDA, ICA, SVM&SVD, ORL Database.