(Georgia Institute of Technology, 2020-05)
Xu, Ruihan
Virtual reality (VR) developed as an immersive communication tool, where users are able to interact with each other through virtual avatars. Among the current VR applications with human avatars, most of them emphasize tracking the upper face, like eye tracking. However, the lower face contains the largest range of motions on the face, yet is usually hard for commercial VR headset to capture. In this work, we implemented two VR communication systems that are focusing on lip and jaw motion: 1) facial landmark tracking, and 2) audio-driven lip synchronization. We then conducted users studies to compare these two systems and therefore explore the role of lip and jaw motions when VR is used as a communication medium.