Integrated Object Recognition for Service Robots Based on Ontology with the Interactive Support from Humans

2014 Fiscal Year Final Research Report

• Project/Area Number: 23300065
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Perception information processing/Intelligent robotics
• Research Institution: Saitama University
• Principal Investigator: KUNO Yoshinori 埼玉大学, 理工学研究科, 教授 (10252595)
• Co-Investigator(Kenkyū-buntansha): KOBAYASHI Yoshinori 埼玉大学, 大学院理工学研究科, 助教 (20466692) ; KACHI Daisuke 埼玉大学, 教養学部, 教授 (50251145)
• Project Period (FY): 2011-04-01 – 2015-03-31
• Keywords: コンピュータビジョン / 知能ロボット / ヒューマンインタフェース / オントロジー / 物体認識 / 対話システム

Outline of Final Research Achievements

We have been developing a helper robot that is able to fetch objects requested by users. Such a robot must recognize the desired object(s) in order to carry out its tasks. However, it is difficult for a system to recognize objects autonomously without fail under various real-world conditions. To address this problem, we have proposed an integrated object recognition system. The system first attempts autonomous object recognition. If it fails, it switches to an interactive object recognition mode where the system asks the user to verbally provide information about the object that it is unable to detect autonomously. However, natural language descriptions from humans can be complex. Such descriptions may not even have a one-to-one correspondence to the physical attributes of objects. We have constructed an ontology for representing such complex relationships between human descriptions and physical attributes and developed a method to recognize objects based on this ontology.

Free Research Field

知能ロボット