Purchase this article with an account.
Jacqueline Snow, Taylor Coleman, Melvyn Goodale; Real-world size improves object recognition in visual form agnosia. Journal of Vision 2014;14(10):188. doi: https://doi.org/10.1167/14.10.188.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Patients with visual form agnosia have difficulty recognizing pictures of objects but can show improved recognition for real-world exemplars – a phenomenon known as the 'Real Object Advantage'. For most everyday objects, visual size corresponds to real-world size, whereas in standard pictorial tests of object recognition stimuli can often be orders of magnitude smaller than their real-world size. Here we investigated the influence of real-world size on visual object recognition in a patient with visual form agnosia (DF). Recognition performance and response times were measured for real objects that were either 'real size' (e.g., banana) or 'toy size' (e.g., minature aeroplane). Because DF can utilize color and surface texture cues to facilitate recognition, all stimuli in the current study were painted white and had little visible surface texture. We compared DF's recognition performance in a number of different viewing conditions: monocular vs. binocular viewing, and free vs. fixed head position. We also examined DF's ability to recognize photographs of the same set of objects (under binocular free viewing) which were closely matched to the real objects for visual size, viewpoint, and illumination. DF's recognition performance in all viewing conditions (except monocular with fixed head position) was better for real 3D objects than for 2D pictures, and she showed small qualitative improvements in recognition performance for binocular over monocular viewing and with free-viewing over fixed head position. DF's recognition accuracy was strongly influenced by real-world size, with better recognition performance for real-world than toy-sized exemplars. Agnosia patients rely largely upon 'top-down' predictions about object identity based on the size of an object, whereas 'bottom-up' visual shape cues provide secondary information about identity.
Meeting abstract presented at VSS 2014
This PDF is available to Subscribers Only