Solving the binding problem and modeling the integration process of multidimensional information based on the no-triplet hypothesis

2017 Fiscal Year Final Research Report

• Project/Area Number: 26590173
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Experimental psychology
• Research Institution: University of Tsukuba
• Principal Investigator: MORITA Masahiko 筑波大学, システム情報系, 教授 (00222349)
• Co-Investigator(Kenkyū-buntansha): 森田 ひろみ 筑波大学, 図書館情報メディア系, 准教授 (00359580)
• Research Collaborator: ISHIZAKI Takuya ; FUJII Yumiko ; FURUTATE Mizuki
• Project Period (FY): 2014-04-01 – 2018-03-31
• Keywords: 視覚特徴統合 / バインディング問題 / 対属性モデル / 刺激反応学習

Outline of Final Research Achievements

How different kinds of information or attributes are integrated in the brain is one of the most fundamental open questions, which is computationally difficult because it involves the combinatory explosion problem. We hypothesized that only pairs of attributes are integrated at a time, and that unified representations of three or more attributes are not formed except for a limited number of very familiar cases. In this study, we obtained strong evidence for this hypothesis, and developed a paired-attribute model based on the hypothesis. We also showed that this model is useful in some applications.

Free Research Field

神経情報処理