Users' head movement obtained from avatars in social VR applications
Chloë Farr, Yadong Liu, Bryan Gick · The Journal of the Acoustical Society of America · 2020
Virtual Reality has been endorsed as a highly beneficial tool for face and hand-head gesture research [Sidenmark and Gellersen (2019), ACM Trans. Computer-Human Interaction 27(4)]. Studies have shown [i.e., Xu, X. et al., J. Biomech. 48(4), 721–724 (2015)] that avatars accurately simulate user movements and allow researchers to obtain users' movements remotely without in-person contact or external recordings taken by the users. This study examines head movements of avatars in three popular social VR applications (Altspace, Oculus Home, and vTime XR) using Openface 2.0 [Baltrusaitis et al., in 13th IEEE International Conference on Automatic Face & Gesture Recognition (FG 2018) (2018), pp. 59–66], software that measures head movements in a video. In each of these three platforms, we recorded a user making a sequence of head movements, rotating side-to-side and up-and-down to their maximum extension. Pilot results show that avatar movements in all three apps could be tracked by OpenFace 2.0. Further analysis will include which application's avatars best represent the user's real-time movements. This study provides insight into speech behaviour research and whether user movements are analysable via an avatar, without in-person contact or lab facilities under the new COVID-19 regulations.