Abstract
Visual information from a face can be broadly divided into either surface-based texture patterns or shape-based cues derived from feature edges. Contrast negation disrupts the surface texture of faces, but leaves shape-based information largely intact. Our aim was to use contrast negation to determine the relative contribution of these different types of facial information to the perception and neural representation of identity and expression. In Experiment 1, we asked how judgements of expression and identity are affected by contrast negation. Participants had to judge either the identity or the expression of two sequentially presented gray-scale faces in 4 conditions: (1) same-expression, same-identity; (2) same-expression, different-identity; (3) different-expression, same-identity; (4) different-expression, different-identity. The faces in each condition were either both positive-contrast, both negative-contrast, or mixed positive- and negative-contrast. Contrast negation had a significant effect on judgements of facial identity, but only a marginal affect on judgements of expression. This shows that judgements of facial identity are more sensitive to surface texture information. Experiment 2 used fMRI to ask whether the neural representations in different face-selective regions are sensitive to surface-based texture patterns. Twenty-five participants viewed blocks of faces that had either the same identity and expression or different identities and expressions. The faces in each block were either (1) all positive contrast, (2) all negative contrast or (3) alternated between positive and negative contrast. We found significant adaptation (different > same) to positive-contrast, but not to negative-contrast or positive/negative-contrast faces in the FFA. However, significant adaptation to all face conditions was evident in the pSTS. These results suggest that neuronal representation in the FFA is sensitive to surface-based cues that are important for the perception of facial identity. In contrast, the neural representation in the pSTS is dependent upon shape-based cues that are important for the perception of facial expression.
Meeting abstract presented at VSS 2013