Learning gait parameters for locomotion in virtual reality systems

Abstract

Mechanical repositioning is a locomotion technique that uses a mechanical device (i.e. locomotion interface), such as treadmills and pedaling devices, to cancel the displacement of a user for walking on the spot. This technique is especially useful for virtual reality (VR) systems that use large-scale projective displays for visualization. In this paper, we present a machine learning approach for developing a mechanical repositioning technique based on a 1-D treadmill for interacting with a unique new large-scale projective display, named as the Wide-Field Immersive Stereoscopic Environment (WISE). We also assessed the usability of the proposed approach through a novel user study that asked participants to pursue a rolling ball at variable speed in a virtual scene. Our results show that participants differ in their ability to carry out the task. We provide an explanation for the variable performance of the participants based on the locomotion technique.