NEW MULTIPLE CLASSIFIER SYSTEMS FOR FACE AUTHENTIFICATION

Abstract

In this paper a multiple classifier systems for face verification is proposed based on the study of scores fusion for four face authentication systems. Extraction features is realized by the Gabor wavelets phases, Principal Component Analysis (PCA) with the Enhanced Fisher linear discriminant Model (EFM) are used as a method of reducing data space. For the study of fusion of scores we used two approaches, the first based on the classification of scores using Fisher statistical method, Support Vector Machine (SVM) and artificial neural networks (MLP) and the second is based on combinations of scores by the weighted sum and fuzzy logic.