论文信息 - Cognitive Modeling of Human-Robot Interaction Estimating Other’s Internal State

Cognitive Modeling of Human-Robot Interaction Estimating Other’s Internal State

Various types of action decision strategy are used to realize a smooth interaction with others. We can estimate intention of others and determine own action accordingly, or can induce intention and actions of others’ as we intend. In this study, a computational model of the action decision has been constructed based on an intention estimation of others and is evaluated its effectiveness by ‘a robot-child play’ experiment. For the model construction, the human adult behavior in the game play with child has been analyzed.. As a result, a necessity of “Meta-strategy” such as a choice of action decision strategy is suggested.

Kasumi Abe | Takashi Omori | Takayuki Nagai | Ayami Yokoyama

[1] Takashi Omori,et al. Computational Modeling of Human-Robot Interaction Based on Active Intention Estimation , 2007, ICONIP.

[2] Tomoaki Nakamura,et al. Learning novel objects using out-of-vocabulary word segmentation and object extraction for home assistant robots , 2010, 2010 IEEE International Conference on Robotics and Automation.

[3] Takashi Omori,et al. Modeling of action decision process based on intention estimation , 2008 .

[4] Kazuyuki Aihara,et al. Multi-agent reinforcement learning algorithm to handle beliefs of other agents' policies and embedded beliefs , 2006, AAMAS '06.