Simulating anthropomorphic upper body actions in virtual reality using head and hand motion data

Abstract

The use of self avatars in virtual reality (VR) can bring users a stronger sense of presence and produce a more compelling experience by providing additional visual feedback during interactions. Avatars also become increasingly more relevant in VR as they provide a user with an identity for social interactions in multi-user settings. However, with current consumer VR setups that include only a head mounted display and hand controllers, implementation of self avatars are generally limited in the ability to mimic actions performed in the real world. Our work explores the idea of simulating a wide range of upper body motions using motion and positional data from only the head and hand motion data. We present a method to differentiate head and hip motions using information from captured motion data and applying corresponding changes to a virtual avatar. We discuss our approach and initial results.