Natural Listening for a Humanoid Robot

概要

In human-robot interactions envisioned in the near future, the robot should be able to use both verbal and nonverbal modalities to communicte with humans in a natural way. This project presents our natual nonverbal listening system for humanoid robots applied to the Robovie II robot. The system uses motion captured data, and sound information to allow the robot to follow the human during an explanation session resulting in a suspension of disblieve. The system was built using the EICA architecture develped to integrate natural interaction with autonomy, and is trained using natural human-human explanations.

URL

http://www.ii.ist.i.kyoto-u.ac.jp/~yasser/GCOE/Listener/

産業界への展開例・適用分野

Applications of the proposed system include assistant and companion robots that can interact naturally with humans, robots that can work as knowledge media in large factories transfering knowledge about various devices and tasks from experts to novice workers and Ambient Assistant Living applications (AAL) in which robots can serve as an assistant for senior citizens.

展示責任者

氏名: Yasser Mohammad

所属: 情報学研究科 知能情報学専攻 知能情報ソフトウェア講座　知能情報応用論分野（西田・角研）

役職: 博士課程後期２年