Uncovering the visual features relevant to human visual scene processing Restricted; Files & ToC

Abstract

Decades of neuroimaging studies have revealed a dedicated cortical network that is selectively involved in human visual places (or “scenes”) processing, including the parahippocampal place area (PPA) (Epstein & Kanwisher, 1998), the retrosplenial complex (RSC) (Maguire, 2001) and the occipital place area (OPA) (Dilks et al., 2013). But what exactly characterizes visual inputs as a “scene” (versus an object, for example) and thereby selectively engages the cortical scene processing system? Here, I propose that visual scenes can be characterized by a set of visual features that are common and unique to visual scenes, and identify two of such features. In Paper 1, I hypothesized that since we are always inside a scene, and interact with the outside of an object, concavity is a diagnostic feature of visual scenes. Consistent with my hypothesis, I found that two scene-selective cortical regions—PPA and OPA—responded selectively to visual cues of concavity, in scene and even object stimuli; moreover, participants also behaviorally judged concave over convex scene and object stimuli to be more like a scene. In Paper 2, using a stimuli-driven approach, I uncovered another visual feature common and unique to visual scenes—vertical luminance gradient (VLG)—and hypothesized that VLG is another diagnostic feature of visual scenes. Consistent with my hypothesis, I found converging neural and behavioral evidence for VLG as a visual feature that humans use for visual scene recognition. In Paper 3, I next asked: Having recognized visual inputs as a “scene”, how then do humans discriminate between different scenes (e.g., a kitchen versus a beach)? Building upon evidence in the existing literature, I hypothesized that color is a visual feature that humans use for visual scene discrimination, and tested whether color is used for visual scene discrimination more specifically, or discrimination of all kinds of visual stimuli more generally (e.g., objects too). I found that participants selectively used color for behavioral visual scene over object discrimination. Moreover, using an existing fMRI database, I also found that color better explained the pattern of response in PPA than an object-selective region (the lateral occipital complex; LOC). Taken together, these findings shed light on the unique properties of visual scenes, and provide insight into how and why visual scenes engage distinct cortical processing in the human visual cortex.

Table of Contents

This table of contents is under embargo until 19 August 2028

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Psychology
Subfield / Discipline	Cognition and Development
Degree	Ph.D.
Submission	Dissertation
Language	English
Research Field	Psychology, Experimental Psychology, Cognitive
关键词	High level vision Scene Vision
Committee Chair / Thesis Advisor	Daniel Dilks, Emory University
Committee Members	Lynne Nygaard, Emory University Joseph Manns, Emory University Phillip Wolff, Emory University Dirk B. Walther, University of Toronto

最新修改

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	File download under embargo until 19 August 2028	2022-05-25 12:20:49 -0400	File download under embargo until 19 August 2028

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions