Evaluation of Superimposed Self-character Based on the Detection of Talkers' Face Angles in Video Communication

摘要

We build upon an embodied video chat system, called E-VChat, in which an avatar is superimposed on the other talker's video images to improve the mutual interaction in remote communications. A previous version of this system used a headset-type motion capture device. In this paper, we propose an advanced E-VChat system that uses image processing to sense the talker's head motion without wearing sensors. Moreover, we confirm the effectiveness of the superimposed avatar for face-to-face communication in an experiment.