体内バーチャルエージェントとのマルチモーダル対話のための音像定位装置の開発

抄録

Many types of companion robots have been studied. We propose a way to interact with a virtual companion agent that is localized inside our bodies. A psychological presence of a companion agent is arranged inside the listener’s body using technology for localizing a sound image. As a first step in the research, we developed a four-channel surround speaker that wraps around the abdomen of the listener. The sound image can move by reproducing the voice of the companion agent and adjusting the amplitude of the four loudspeakers. Furthermore, by using pressure sensors and Peltier elements, the listener can engage in multimodal interactions with the in-body virtual agent using tactile senses in addition to voice. In this paper, we introduce a way to enhance the in-bodiness of an in-body virtual agent for multimodal interaction.