Situated perception of a partner robot based on neuro-fuzzy computing

This paper discusses the situated perception of a partner robot interacting with a human. The situated perception is based on the mutual cognitive environment discussed in relevance theory. It is very important to share actual environment in order to realize natural communication, because the meaning of natural language and gesture can be translated according to the surrounding environment. Next, we discuss the role of imitation based on gesture, and explain the method for imitative behavior generation. Finally, we show experimental results of situated perception, imitative behavior generation, and behavior coordination through interaction with a human.