Neuropsychological Components of Object Identification
Abstract
Four kinds of investigations of the structure of high-level visual processing were conducted during the past year. First, we carried out case studies of individual brain-damaged patients, finding evidence that curved edges are processed separately from straight edges, that location information sometimes can be used to encode some characteristics of shape, and that a decrease in overall activation level can selectively impair performance on some tasks. Second, we administered a set of 27 tasks to a group of 17 brain-damaged patients; these tasks were designed to assess the efficacy of specific subsystems. We have preliminary evidence that most (89%) of the subsystems double dissociate-suggesting that these subsystems are in fact distinct. Third, we constructed computer simulation models to explore properties of the high- level visual system, and found support for the distinction between subsystems that compute two distinct kinds of spatial relations. Finally, some of the tasks we had developed to study deficits in brain-damaged patients were used to study the visual-spatial abilities of air force pilots; we found that pilots are particularly good at mental rotation and encoding metric distance information. Neuropsychology, Vision, Computational models.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 10, 1992
- Accession Number
- ADA247049
Entities
People
- Stephen M. Kosslyn
Organizations
- Howard University