Performance Improvement and Skill Transfer in Table Tennis Through Training in Virtual Reality

IEEE Trans Vis Comput Graph. 2022 Dec;28(12):4332-4343. doi: 10.1109/TVCG.2021.3086403. Epub 2022 Oct 26.

Abstract

Sports professionals have been increasingly using Virtual Reality (VR) for training and assessment of skill-based sports. Yet fundamental questions about the virtue of VR training for skill-based sports remain unanswered: Can the complex motor skills in these sports be learned in VR? If so, do these skills transfer to the real world? We have developed a VR table tennis system that incorporates customized physics with realistic audio-visual stimuli, haptics, and motion capture to enhance VR immersion and collect information about the player's posture and technique. We have assessed skill acquisition and training transfer by comparing real table tennis performance between a control group (n=7) that received no training and an experimental group (n=8) trained for five sessions in VR. Results show a significant improvement in technique but no significant changes in the number of the returned balls in the experimental group in the real-life retention session. However, no significant differences are found in the control group. Our findings support the notion that complex skills can be learned in VR and that obtained skills can transfer to the real world. This work offers an inexpensive VR table tennis training platform, enabling effective training via real-time motor and ball returning technique feedback.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Computer Graphics
  • Motor Skills
  • Tennis*
  • User-Computer Interface
  • Virtual Reality*