Intel Gifts UTSA $200k to Support Deep Learning Research to Reduce Cybersickness in Virtual Reality

Dr. John Quarles, associate professor of computer science, has received a gift of $200,000 from Intel in support of his research in using deep learning to reduce cybersickness in virtual reality. The title of his project is “Cyberwell: a closed loop system using pipeline of deep learning models and EEG based Brain Computer Interfaces to HMDs to enable personalized, real-time cybersickness reduction.”

Cybersickness is a major barrier to widespread user acceptance of virtual reality (VR) and there are currently minimal solutions to practically address this issue. Common symptoms of cybersickness are similar to motion sickness, including dizziness, disorientation, headache, and nausea, all of which can continue to persist for hours after being exposed to VR. However, the incidence and severity of cybersickness can be highly individualized, can be difficult to predict when onset will occur, and can be affected by VR system aspects (e.g. latency), stimuli within a specific virtual environment, and user interaction behavior. Although there are many known ways of detecting and reducing cybersickness, there are no real-time, closed loop approaches that can automatically predict and reduce cybersickness, which is the focus of this research.

Quarles’ goal is to create a system that enables automatic, closed loop cybersickness reduction. He expects that this will improve user acceptance of VR, improve the performance of VR to offer personalized experience, and the reduction of cybersickness in general will enhance human performance in VR applications. The proposed research will initially focus on investigating how to reduce cybersickness in VR, but the broader vision of this project is that it can be applied to many other application areas within VR/AR/MR, including healthcare, training and education, and entertainment.

For more information on the work Dr. Quarles and his San Antonio Virtual Environments (SAVE) Lab conducts, click here.