VR2ML: A Universal Recording and Machine Learning System for Improving Virtual Reality Experiences
Yuto Yokoyama, Katashi Nagao · 2020 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW) · 2020
We developed software for recording and reusing user actions and associated effects in various VR applications. In particular, we implemented two plugin systems that allow 3D motion data recorded in VR to be transferred to a machine learning module without programming. One is a system that runs on Unity and records actions and events in VR, and it is called “VRec.” The other is a system for using data recorded with VRec for machine learning, called “VR2ML.”