Item

Abstract

Categorization researchers and, more recently, object recognition researchers have proposed that similarity may offer an organizational principle for the representation of
objects in the brain and have demonstrated that such a principle is capable of
supporting the distinct tasks of object recognition and categorization. But what is it that
makes two objects perceptually similar? And what role does sensory modality play in
shaping perceptual similarity? As a first step towards addressing this question, we used
computer graphics modeling techniques to create novel, 3D objects that varied
parametrically in shape and texture. In a psychophysical experiment, we then asked
human subjects to view pairs of objects and rate the similarity between them. We also
generated similarity ratings from a set of standard computer vision techniques (based
on pixel-wise differences, cross-correlations, edge images, and Gabor jets).
Multidimensional scaling (MDS) on the perceptual similarity data revealed that the
perceptual representation of both shape and texture changes is veridical; the twodimensional
MDS configuration of stimuli closely matches the configuration in shape
and texture parameter space. Interestingly, comparisons against computational
measures revealed that simple pixel-based and correlation-based similarities were
closest to visual similarities. In future experiments, we plan to compare haptic
similarity measures to these visual measures in order to examine the role of modality in
mediating the effects of shape and texture changes on perceptual similarities.