Eye tracking performance assessment — for VR/AR headsets and wearables

Summary

The inclusion of eye tracking in VR/AR headsets and other wearables has reached a tipping point in maturity. The technology is an enabler for techniques such as foveated rendering and foveated transport — which contribute to enhanced device performance and lower bandwidth demands. Consequently, eye tracking has become a highly sought-after capability in VR, AR, and other wearable devices.

Developers and ISVs are leveraging eye tracking to create immersive XR experiences with features like eye-controlled menu selection. The insights eye tracking generates enables informed decision making, helping organizations, enterprises, and individuals to solve challenging issues such as brain trauma and cognitive load assessment.

Unfortunately, the performance capabilities of any given eye tracking system can vary greatly. To overcome the challenges such variation presents, we propose a set of assessment metrics together with a methodology for collecting the underlying data points. Our aim is to provide a platform for comparing the performance of different eye tracking systems and how they vary for each user and use case. And to ensure success in the consumer market, our assessment metrics consider population coverage for at least the 95th percentile.