2019 年 21 巻 3 号 p. 293-302
We have developed a speech-driven embodied entrainment computer graphics character called ”InterActor” which generates automatically communicative motions such as nodding from the rhythm of speech, and have demonstrated the effectiveness of the system for supporting embodied communication. However, though conventional InterActor activates embodied interaction, it does not have expressions accompanying time transition such as accumulation of utterances and delimitation of speech. Visualizing user’s speech transition over time would be effective for motivating user’s utterances. In this paper, in order to increase user’s motivation to speech, we develop an utterance transmission / accumulation expression system by using InterActors expressing each function of speaker and listener that expresses the transmission of the utterances from the speaker character to the listener character and the accumulation to the listener character by expansion accompanying nodding. We demonstrate the effectiveness of the system in free conversation experiments. Furthermore, in order to realize the responses that urge more continuous utterances by eliminating difficulty of speaking due to excessive expansion of the listener character, we construct a system that supports delimitation of dialogue by changing listener characters, and demonstrate the effectiveness of the system in free conversation experiments.