Files in this item



application/pdfNG-DISSERTATION-2021.pdf (2MB)
(no description provided)PDF


Title:The parallel nature of early visual processing in visual search
Author(s):Ng, Gavin JP
Director of Research:Lleras, Alejandro; Buetti, Simona
Doctoral Committee Chair(s):Lleras, Alejandro; Buetti, Simona
Doctoral Committee Member(s):Beck, Diane M; Christianson, Kiel; Hummel, John E
Department / Program:Psychology
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):visual search
parallel processing
target contrast signal theory
Abstract:Visual search is a process that is ubiquitous in daily life. Some are relatively easy – looking for a red sock among black socks – while others are extremely difficult – looking for tumors in x-ray images. Classic theories and models of search have long assumed that these easy searches, referred to as efficient search, occur in parallel and without attention and thus should not be affected by factors such as the number of items or the features of the distractors. However, there has been recent evidence that systematic variability exists even inefficient search. This led to the development of the Target Contrast Signal Theory, which proposed a different mechanism for parallel processing in visual search. In contrast to the pre-attentive process of saliency detection and generation proposed by classic theories, Target Contrast Signal Theory proposes that parallel processing in early vision involves an evidence accumulation process that is aimed at rejecting items as non-targets. In this thesis, I present a series of studies that examined the nature of the parallel process in detail. Chapter 2 examines the output of parallel processing and provides evidence in support of the mechanism described in Target Contrast Signal Theory. Chapter 3 examines the fate of the items that have been discarded in parallel processing. Lastly, Chapter 4 examines the influence of scene context on the parallel evidence accumulation process.
Issue Date:2021-04-23
Rights Information:Copyright 2021 Gavin Ng
Date Available in IDEALS:2021-09-17
Date Deposited:2021-05

This item appears in the following Collection(s)

Item Statistics