Purchase this article with an account.
Guy Wallis; A multichannel model of face processing based on self-organizing principles. Journal of Vision 2013;13(9):1115. doi: https://doi.org/10.1167/13.9.1115.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Evidence from developmental, behavioral, and electrophysiological studies suggests that the recognition and representation of faces is subserved by specialist cortical processes and sub-regions, separate from those involved in object recognition. This dichotomy is reflected in a lack of overlap between models of face and object recognition. Models of face processing have traditionally focused on behaviors thought to be unique to face recognition such as the prototype, configural, and holistic effects, and face-specific adaptation after-effects. Currently, the most widely accepted theory of object recognition is based not on the norm-based encoding models of face processing, but rather, on abstract features whose size and complexity increase over a series of competitive, hierarchical stages. Despite the fundamentally piecewise nature of such a model, I explain how effects such as prototyping, configural and holistic coding, are all emergent properties of such a system when one considers the impact of learning. The ideas are backed up by novel behavioral results and simulation of an unsupervised competitive network. The results highlight how the peculiarities of face processing can be seen as an inevitable consequence of our intense exposure to faces.
Meeting abstract presented at VSS 2013
This PDF is available to Subscribers Only