In this paper we study the reliability of the methods of Face Recognition on the ground of the precision of Face Detection. We have made an extensive study on methods for feature extraction, feature transformation and classification in the problem of Face Recognition. In this work, we report an experimental comparison of several learning systems coupled with different feature representations. Extensive experiments carried out on the ORL database and YALE-B database, which are the most common benchmarks in this area, prove the advantage of a new approach based on an ensemble of classifiers when compared with other well-known techniques. Our experiments show that combining different face recognition approaches we obtain a very low equal error rate.
Combining classifiers to obtain a reliable method for face recognition
LUMINI, ALESSANDRA;NANNI, LORIS
2005
Abstract
In this paper we study the reliability of the methods of Face Recognition on the ground of the precision of Face Detection. We have made an extensive study on methods for feature extraction, feature transformation and classification in the problem of Face Recognition. In this work, we report an experimental comparison of several learning systems coupled with different feature representations. Extensive experiments carried out on the ORL database and YALE-B database, which are the most common benchmarks in this area, prove the advantage of a new approach based on an ensemble of classifiers when compared with other well-known techniques. Our experiments show that combining different face recognition approaches we obtain a very low equal error rate.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
