Categorization and Expertise in Human Visual Cognition

点击翻译按钮获取中文摘要

9615819 TARR This research will investigate visual recognition mechanisms employed by human observers across different levels of categorization and expertise, with particular reference to the difference between faces and non-face objects. Earlier studies making claims for face-specific recognition mechanisms include a number of experimental confounds. Addressing these concerns will first require the use of computer graphics to create carefully controlled three-dimensional objects for presentation in psychophysical, neuropsychological, and functional neuroimaging experiments. These objects will include a set of non-face control objects that will combine several properties of faces (such as a fixed configuration of parts and homogeneity across the class) but none of their superficial characteristics. A first set of psychophysical experiments will use these novel stimuli, as well as familiar classes of objects, to assess which of the several "face-specific" behavioral effects described in the literature can be obtained for a) individual as opposed to more categorical judgments for the same objects and, b) with experts as opposed to novice participants. New expertise training procedures will provide methods for manipulating participants' experience with the novel object class as well as a better understanding of the reorganization of recognition mechanisms that may occur during expertise acquisition. A second set of experiments will test brain-lesioned patients traditionally classified as having a face-specific disorder in neuropsychological experiments with the same object class to test whether face-recognition deficits may be better described as an impairment of exemplar-level processing mechanisms. Testing of these patients will be performed in a manner closer to what is standard in the testing of normal participants, by improving the control of object presentation and of response measurements, in particular using reaction times and signal detection measure s. Other experiments will use methods developed for studying the creation of expertise in normals to explore the plasticity of the visual system of patients with recognition deficits. A third set of experiments will study the neural substrate of exemplar vs. categorical level visual judgments using functional magnetic resonance imaging (fMRI). While the importance of comparing face processing to exemplar-level processing of non-face objects has been acknowledged, few studies have compared recognition judgments at different levels of abstraction for the same objects, in particular, under conditions where participants begin as novices and then gain experience with an object class. Longitudinal fMRI studies will assess the neural changes underlying expertise acquisition at multiple levels of categorization. Overall, these experiments should shed some light on the issue of face-specificity in the visual system, provide data about the role of categorization level and expertise for object recognition, explore rehabilitation possibilities for patients with visual form recognition deficits, and extend our understanding of the plasticity of the human visual system. ***