The style in which a robot moves, expressed through its gait or locomotion, can convey effective messages to people. For example, a robot could move aggressively
in reaction to a person’s actions, or alternatively react using a set of careful, submissive movements. Designing, implementing and programming robotic interfaces
that react to users’ actions with properly styled movements can be a difficult, daunting, and time consuming technical task. On the other hand, most people can easily
perform such stylistic tasks and movements, for example, through acting them out. Following this observation, we propose to enable people to use their existing teaching
skills to directly demonstrate to robots, via in-situ acting, a desired style of interaction. In this paper we present an initial style-by-demonstration (SBD) proof-of-concept
of our approach, allowing people to teach a robot specific, interactive locomotion styles by providing a demonstration. We present a broomstick-robot interface for directly
demonstrating locomotion style to a collocated robot, and a design critique evaluation by experienced programmers that compares our SBD approach to traditional programming methods.
Researchers: James E. Young, Ehud Sharlin, Takeo Igarashi, Kentaro Ishii
- J. E. Young, K. Ishii, T. Igarashi, E. Sharlin, “Style by Demonstration: Teaching Interactive Movement Style to Robots”, Proceedings of the ACM International Conference on Intelligent User Interfaces (IUI 2012). Lisbon, Portugal (best paper award).
- J. E. Young, E. Sharlin, T. Igarashi, “Teaching Robots Style:Designing and Evaluating Style-by-Demonstration for Interactive Robot Locomotion”, HCI Journal. (in press)
- J. Young, K. Ishii, T. Igarashi and E. Sharlin, “Style by Demonstration: Using Broomsticks and Tangibles to Show Robots How to Follow People” Departmental Technical Report 2010-975-24, October 13, 2010, Department of Computer Science, University of Calgary, AB, Canada, 2010 .
- J. Young, K. Ishii, T. Igarashi and E. Sharlin, “Style-by-Demonstration: Using Broomsticks and Tangibles to Show Robots How to Follow People”, Departmental Technical Report 2009-940-19, November 3, 2009, Department of Computer Science, University of Calgary, AB, Canada, 2009.