Please use this identifier to cite or link to this item: http://scholarbank.nus.edu.sg/handle/10635/107375
Title: INVESTIGATING THE IMPORTANCE OF OBJECT AND SEMANTICS IN GAZE DEPLOYMENT : A DATASET AND ANALYSES.
Authors: XU JUAN
Keywords: computational model, dataset, object saliency, saliency attribute, semantic saliency, visual saliency
Issue Date: 24-Jul-2014
Citation: XU JUAN (2014-07-24). INVESTIGATING THE IMPORTANCE OF OBJECT AND SEMANTICS IN GAZE DEPLOYMENT : A DATASET AND ANALYSES.. ScholarBank@NUS Repository.
Abstract: MOST PREVIOUS VISUAL ATTENTION WERE BOTTOM-UP MODELS BASED ON PIXEL-LEVEL FEATURES. THEY PREDICT WHERE TO LOOK IN A NATURAL IMAGE BY COMPUTING A "SALIENCY MAP" THAT HIGHLIGHTS THE REGIONS STANDING OUT OF THE SCENE. HOWEVER, RECENT EVIDENCE SUGGESTS THAT HUMAN VISUAL ATTENTION IS ATTRACTED BY INTERESTING OBJECTS IN THE SCENE, BUT NOT ONLY THE PIXEL-LEVEL FEATURES. OBJECTS AND SEMANTICS MAY PLAY MORE IMPORTANT ROLES IN ATTENTIONAL SELECTION. THEREFORE, TO BRIDGE THE SEMANTIC GAP BETWEEN THE PREDICTIVE POWER OF COMPUTATIONAL SALIENCY MODELS AND HUMAN BEHAVIOR, WE PROPOSE A NEW SALIENCY ARCHITECTURE THAT INCORPORATES INFORMATION AT THREE LAYERS: PIXEL-LEVEL IMAGE ATTRIBUTES, OBJECT-LEVEL ATTRIBUTES, AND SEMANTIC-LEVEL ATTRIBUTES. OBJECT- AND SEMANTIC-LEVEL INFORMATION IS FREQUENTLY IGNORED, OR ONLY A FEW SAMPLE OBJECT CATEGORIES ARE DISCUSSED WHERE SCALING TO A LARGE NUMBER OF OBJECT CATEGORIES IS NOT FEASIBLE NOR NEURALLY PLAUSIBLE. TO ADDRESS THIS PROBLEM, THIS WORK CONSTRUCTS A PRINCIPL
URI: http://scholarbank.nus.edu.sg/handle/10635/107375
Appears in Collections:Master's Theses (Open)

Show full item record
Files in This Item:
File Description SizeFormatAccess SettingsVersion 
thesis.pdf14.39 MBAdobe PDF

OPEN

NoneView/Download

Page view(s)

57
checked on Sep 21, 2018

Download(s)

20
checked on Sep 21, 2018

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.