Study on Creativity of Robot Inspired from Subjective Process of Human

2017 Fiscal Year Final Research Report

• Project/Area Number: 16K12509
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Kansei informatics
• Research Institution: Tokyo University of Agriculture and Technology
• Principal Investigator: Mizuuchi Ikuo 東京農工大学, 工学(系)研究科(研究院), 准教授 (60359651)
• Research Collaborator: HAYASHI Kotaro 東京農工大学, 大学院工学府, 特任助教 (80728345)
• Project Period (FY): 2016-04-01 – 2018-03-31
• Keywords: ロボット / 人とロボットの関わり / ヒューマノイド / 表情認識 / 心的状態 / 創造

Outline of Final Research Achievements

The aim of this study is to utilize humans' subjective process (such as feeling beauty of arts or sports, feeling hospitality, having interest in what we do not know, etc.) for motion/behavior generations and/or creative activities by a robot or an artificial system. We have proposed to construct an estimator of subjective evaluation of a human, based on the machine learning technology. We collected various real data of human's subjective evaluation and used them as supervisor data for learning.
Although human's subjective evaluation process has intrinsic fluctuation, we have succeeded to create several systems that select or generate motions or behaviors of robots or installations, or novel paintings. These results have been presented in international and domestic conferences and exhibitions. We have also started new studies on manipulating an impression on a non-humanoid robot, automatic generation of skillful sensory-motor robot motions.

Free Research Field

Robotics