Non-linear dynamic models of search behavior: Eye movement and haptic search

2015 Fiscal Year Final Research Report

• Project/Area Number: 26540065
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Cognitive science
• Research Institution: Kyoto University
• Principal Investigator: Saiki Jun 京都大学, 人間・環境学研究科(研究院), 教授 (60283470)
• Co-Investigator(Kenkyū-buntansha): UEDA Yoshiyuki 京都大学, こころの未来研究センター, 特定助教 (80582494)
• Co-Investigator(Renkei-kenkyūsha): SAKAGAMI Masaaki 京都大学, 大学院人間・環境学研究科, 教授 (70202083)
• Project Period (FY): 2014-04-01 – 2016-03-31
• Keywords: 視覚探索 / 眼球運動 / 非線形力学 / 顕著性マップ / アンサンブル情報 / 自然画像

Outline of Final Research Achievements

Using eye movement data in visual search, we investigated non-linear dynamics in search behavior and examined the limitation of saliency map models. Eye movement analysis in visual search tasks revealed no significant differences in non-linearity of eye movement dynamics between natural scenes with structural properties and noise images without them. However, in addition to saliency, a vanishing point of the scene attracts eye movement, indicating a limitation of saliency map models. To examine the role of global statistical information of an image (ensemble information) in visual search, we devised a task judging the number of targets, which manipulates both targets’ saliency and their numbers. The results showed that ensemble information as well as saliency contributes to eye movements.

Free Research Field

認知神経科学