Visual pattern classification by neural fields
Résumé
The recognition of visual motion patterns such as walking, fighting and face gestures among others, is remarkably efficient in humans and many other species. Experiments have already given some clues about the nature of the internal mechanisms of recognition. These experiments are based on point-light stimuli in psychophysics, electro-physiological data and functional imaging techniques. In this work, we study some of the identified properties and we propose to model them by means of asymmetric neural fields.
Domaines
Réseau de neurones [cs.NE]
Origine : Fichiers produits par l'(les) auteur(s)