Plasticity of face perception in the human visual system

Abstract

When our surrounding perceptual environment changes, our sensory systems change in tandem to help us process information efficiently and improve perception. In the visual system, adaptation and perceptual learning are examples of this plasticity. This dissertation presents three studies investigating how adaptation and perceptual learning improve our ability to see faces. Study 1 explored whether higher-level adaptation to faces can help overcome illusory percepts caused by adaptation at lower-level visual stages. We found that the visual system can adapt to distorted faces caused by adaptation at lower-level visual areas through repeated viewing; we termed this phenomenon “meta- adaptation”. Meta-adaptation may be a general strategy to correct negative consequences, e.g. illusory percepts, of lower-level adaptation. Study 2 investigated whether perceptual learning could help observers form mental representations of groups of faces. We showed that 3 days of perceptual learning made faces that initially appeared similar more dissimilar; clustering analysis further suggested that training strengthened group-level representations for the trained faces. In study 3, we more directly tested whether perceptual learning could form group-level representations, by measuring transfer of face adaptation effects between two trained groups of faces. Results showed learning caused a decrease in transfer of adaptation between the trained groups, suggesting that after perceptual learning observers were able to develop more distinct representations for each group. Together, these findings help us better understand how facei perception can be influenced by past viewing history and training; the concepts of meta-adaptation and group-level representations could generalize to other domains of visual perception, and other sensory modalities