Kitagishi, YukiTanaka, YukiYonezawa, TomokoKulik, Alexander and Sra, Misha and Kim, Kangsoo and Seo, Byung-Kuk2020-12-012020-12-012020978-3-03868-112-01727-530Xhttps://doi.org/10.2312/egve.20201276https://diglib.eg.org:443/handle/10.2312/egve20201276In this paper, we propose an augmented reality agent separated from the lecturer as his/her avatar to promote individual communication between the lecturer and the audience in parallel with one-to-many communication such as lectures. The proposed agent is expressed by 1-1) a shadow and 1-2) its trajectory drawn by projection mapping to the floor, as well as 2) localized footsteps using the parametric speaker. The agent approaches a particular audience member who seems to need to talk with the lecturer individually. To emphasize that the agent is the lecturer's avatar, the proposed system draws the trajectory from the lecturer's feet to the agent's current position and shakes the shadow's tail like will-o'-the-wisp. During the lecture, the agent approaches the audience member who wants individual communication with the lecturer. After that, the agent talks to the audience member as the lecturer. We expect that the proposed agent will aid in the individual communication between the lecturer and the audience without the audience's hesitation, and will be able to enable multiple and parallel communications with few additional resources required by the lecturer.Human centered computingMixed / augmented realityApplied computingInteractive learning environments10.2312/egve.2020127615-16