Using a multiclass novelty classifier for face recognition

Face recognition, one of the most explored themes in biometry, is used in a wide range of applications: access control, forensic detection, surveillance and monitoring system, robotic and human machine interaction. In this paper, a new classifier is proposed for face recognition. The performance of this new classifier is compared with the performance of the KNN classifier. The face image database used was the ORL. For feature extractions the following methods were employed: PCA, 2DPCA and (2D)2PCA. The performance tests of both classifiers were done both in verification and identification mode. In identification mode, the recognition rate with the leave-one-out strategy is equal to 100% with PCA, 2DPCA and (2D)2PCA. In the verification mode, the recognition rate is 100% with PCA and 2DPCA and 97.5% for (2D)2PCA. For the half-half strategy, the best recognition rate in the identification mode was obtained with (2D)2PCA, 98.5%, and in the verification mode, with PCA, 88%.