ウ　ジンソク   WOO JINSEOK   禹　珍碩    所属   工学部 機械工学科    職種   専任講師
言語種別	英語
発行・発表の年月	2013/09
形態種別	学術論文
査読	査読あり
標題	Recognition of Indoor Environment by Robot Partner Using Conversation
執筆形態	共著
掲載誌名	Journal of Advanced Computational Intelligence and Intelligent Informatics (JACIII)
掲載区分	国外
出版社・発行元	Fuji Technology Press Ltd.
巻・号・頁	17(5),pp.753-760
総ページ数	8
担当区分	筆頭著者,責任著者
著者・共著者	Jinseok Woo, Naoyuki Kubota
概要	To support daily life before performing an action, a robot partner must perceive an unknown environment. Much research has been done from various viewpoints on self-localization estimation and environment perception. In our research, the robot partner performs self-localization and environment recognition using Simultaneous Localization and Mapping for self-localization estimation and map building. In this paper, we propose a method for recognizing indoor environments by robot partners based on conversations with human beings. Information acquired from maps is identified in order to share the meaning with human beings after the required interpretation. In this paper, we therefore propose a method for recognizing environmental maps by labeling these maps based on symbolic information developed through conversation with human beings. The proposed method is composed of four parts. First, the robot partner applies a steady-state genetic algorithm for self-localization estimation. Second, we use a map building algorithm for expressing the topological map. Third, conversation with human beings is performed for acquiring symbolic information in order to recognize object and position locations through the map. Fourth, we perform experiments and discuss the effectiveness of the proposed technique.
外部リンクURL	https://www.fujipress.jp/jaciii/jc/jacii001700050753/