Apprentissage non supervisé pour la reconnaissance de visage

Abstract

In recent years, interest in biometrics has increased. Facial recognition as a basic biometric technology has become increasingly important in research, given its non-intrusive and contactless nature. But despite the many methods and methods that have been proposed to solve the problem of human facial recognition, it remains a very difficult problem, due to the fact that the faces of different people usually have the same shape and vary due to different lighting, situations and facial expressions. As part of our work, we are interested in the study of the facial method through the development of the facial recognition system. The automatic facial recognition process consists of two steps: extracting and classifying the distinctive elements. This is why we will use the PCA (Principal Component Analysis) method, which uses the reduction property in terms of volume and storage as small radii in matrices for a later comparison.