Abstract
Background: High-level aftereffects can be used to explore the nature of object representations in the human visual system. Whether increased expertise for one object category leads to changes in aftereffects for that category is uncertain. Also, given the potential for competition between representations of different objects, expertise for one category may lead to changes in aftereffects for other objects. Hypothesis: We examined first whether expertise for a category of objects like cars or birds is associated with better discrimination with morphed images of the expert category, or an increase in aftereffect magnitude for that category. Second we examined if discrimination or aftereffect magnitude was reduced for other categories. Method: We tested 30 subjects who were bird experts (n=10), car experts (n=10), or non-experts (n=10). All subjects also received independent tests to confirm perceptual expertise for faces, cars and birds. To assess discrimination and aftereffects, we used a perceptual bias adaptation technique to probe identity aftereffects with morphed stimuli of human faces, birds, and cars. Results: First, the ability to discriminate structural changes differed between the groups only for birds, where the bird experts showed superior discrimination compared to car experts and non-experts. Second, the magnitude of the adaptation for each of the three object categories did not vary with subject group. Finally, at an individual subject level, there was a correlation between the magnitude of aftereffects for faces and birds. Conclusion: These results provide some support for sharpened discrimination as a consequence of expertise, mainly for birds. However, they do not suggest that discriminative ability for non-expert categories suffers as a consequence of perceptual expertise in another category: in fact, there was a positive correlation between the two natural-object categories, faces and birds. Aftereffect magnitude does not appear to be altered as a consequence of perceptual expertise.
Meeting abstract presented at VSS 2014