VR Dodge-ball: Application of Real-time Gesture Detection from Wearables to ExerGaming

Abstract

Wearable technologies are applied to various areas nowadays, including health care and physical training. In previous work, we developed ExerSense, an algorithm to segment, classify, and count different predefined motions in real-time using correlation of acceleration. Using ExerSense, we developed a Virtual Reality (VR) Dodge-ball game. A watch-like wearable device is mounted to the right wrist and connected wireless to a head-mounted display (HMD). ExerSense enables to recognize real-time ball throwing gesture. Dodging opponent's ball in 3-Dimensional virtual space is detected using the sensors embedded in the HMD. We found that wearable devices have an advantage for virtual exergames against the camera-based approaches: there is no space and place limitation. Additionally, the study shows that exergames could be implemented with only general usage devices.