Neural Representation of Faces in Human Visual Cortex: the Roles of Attention, Emotion, and Viewpoint

Abstract

Faces constitute a special class of visual stimuli not only because we possess expert visual skills and specialized brain areas to recognize them, but also because we can extract a rich set of socially and affectively important information from them in a seemingly effortless manner. Abundant research conducted in cognitive psychology, neuroscience, and clinical neuropsychology has provided an elaborate model of the complex functional architecture underlying these different aspects of face processing, each presumably associated with specific neural substrates that are interconnected all together within a large-scale distributed network (Grüsser and Landis 1991). Thus, many influential neurocognitive models have proposed that face recognition may proceed along a series of distinct stages organized in a hierarchical stream of processing (Bruce and Young 1986; Haxby et al. 2000), from low-level visual analysis subserving the detection and organization of facial features, up to higher-level processes allowing the storage and retrieval of personal information and other associative functions (Fig. 1a). Furthermore, some dissociations in recognition performance in healthy subjects, as well as neuropsychological deficits observed in patients with focal brain lesions, have led to the idea that different processing pathways might be responsible for extracting identity-related information versus other facial features related to emotional expression, eye gaze direction, or speech lip motion, and that such pathways might operate in parallel (Bruce and Young 1986; Grüsser and Landis 1991).