Representations of Shape in Object Recognition and Long-Term Visual Memory
Abstract
A variety of studies examining the mechanisms and representations underlying human object recognition have been conducted. One track has investigated the role of view-based object representations in perception and recognition. Results indicate that certain classes of viewpoint-dependent features may be used to define boundaries between characteristic views of objects. A second track has investigated the interaction between orientation- dependent and orientation-independent recognition mechanisms. Results here indicate that humans learn both object-based, orientation-independent and view- based, orientation-dependent representations regardless of the initial learning context. Other results indicate that task conditions mediate whether structural descriptions or episodic representations of objects are used in performing an implicit memory task. Finally, a third track has investigated the nature of spatial relations between objects, as well as the relationship between perceptual and lexical representations of spatial relations. Results indicate that spatial prepositions (e.g., above , left ) encode the relationship between figural and reference objects as a gradient that decreases with distance from the qualitative or veridical position. Moreover, results indicate that this may in part be a lexical effect, in that stronger qualitative effects are found when subjects have lexically encoded the relationship - although further results indicate that qualitative gradients are present in purely perceptual judgments.. .. Object Representation, Object Recognition, Visual Cognition.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 11, 1993
- Accession Number
- ADA264342
Entities
People
- Michael J. Tarr
Organizations
- Yale University