Integration of Intelligence for Speech Act and Motion Act based on Mutual Conversion between Sensorimotor Patterns and Symbols

语言行为和动作行为的智能集成是智能机器人与人类合作的重要功能。针对运动行为的符号化问题，提出了一种抽象感觉运动模式的相空间构造方法。此外，还建立了评价行为和言语策略合理性的随机模型。将两种方法相结合，提出了一种同时控制语言行为和动作行为的结构。

Integration of intelligence for speech act and motion act is an important function for intelligent robots that cooperate with human beings. A phase space construction method to abstract sensorimotor patterns is developed for symbolization of motion act. Additionally, stochastic model that evaluate plausibility of behavior and speech strategy is developed. With integration of the two methods, an architecture that control both of speech act and motion act is introduced.