Texture and object representation for human visual perception
Abstract/Contents
- Abstract
- Human visual object perception occurs in service of behavior. The behaviors which rely on object vision are numerous and varied, from object categorization to visual search to face recognition to novel pattern discrimination. The format of brain representations which underlie visual perception must therefore be sufficiently generalizable as to support this wide variety of behaviors. The visual cortex of humans, especially the ventral visual cortex, contains neural populations which encode information about objects. For example, neurons in primary visual cortex (V1) respond more to lines of some orientations than others, neurons in area V2 respond more to naturalistic textures than to noise patterns, and neurons in the fusiform gyrus respond more to faces than to other objects. Despite decades of research characterizing such neural selectivities, much remains to be understood about how these visual feature representations give rise to perception. In this dissertation, I aim to characterize the format of visual object and texture representations in human ventral visual cortex as well as the downstream cortical computations which select and integrate such representations in support of behavior. I present evidence that human visual cortex encodes information about objects but does not directly support object perception. Rather than explicitly encoding a fixed set of familiar stimuli, visual cortex provides a configurable basis of complex texture-like feature representations. Further, I show that attention is necessary to transform these texture-like representations into a format better suited to support object perception. Finally, I demonstrate that depending on task demands, visual perception must be able to flexibly access feature representations from different visual cortical regions. I conclude by discussing the implications of this work on future modeling efforts for computer vision and cognitive neuroscience. Taken in sum, this dissertation suggests that human visual cortex extracts a complex basis set of features which describe the visual world, and attention and task engagement select and integrate those visual features to support perception.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2022; ©2022 |
| Publication date | 2022; 2022 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Jagadeesh, Akshay Vivek |
|---|---|
| Degree supervisor | Gardner, Justin, 1971- |
| Thesis advisor | Gardner, Justin, 1971- |
| Thesis advisor | Grill-Spector, Kalanit |
| Thesis advisor | Norcia, Anthony Matthew |
| Thesis advisor | Yamins, Daniel |
| Degree committee member | Grill-Spector, Kalanit |
| Degree committee member | Norcia, Anthony Matthew |
| Degree committee member | Yamins, Daniel |
| Associated with | Stanford University, Department of Psychology |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Akshay Vivek Jagadeesh. |
|---|---|
| Note | Submitted to the Department of Psychology. |
| Thesis | Thesis Ph.D. Stanford University 2022. |
| Location | https://purl.stanford.edu/kd735vw1621 |
Access conditions
- Copyright
- © 2022 by Akshay Vivek Jagadeesh
Also listed in
Loading usage metrics...