Abstract

Within Artificial Intelligence, face recognition is an important topic, being related to a large number of practical applications. In particular, in the last years, automatic face recognition systems have grown up, especially for security and surveillance goals. However, it is difficult to introduce an efficient computational approach without a psychological and a physiological knowledge about visual perception. To this aim, this paper contains a short review of cognitive and neuropsychological models of processing underlying face perception. These models, relying on actual knowledge about brain anatomical structures and physiological processes, try to clarify how the brain perceives. We also take into consideration some neural network models that, to some extent, have been designed on the basis of this knowledge. Starting from acquisitions deriving from years of studies about human face recognition could be useful to design new automatic face recognition systems and to improve the performances of the already existing ones.