Weighting VR QoE Factors into a Single Metric

Determine a principled method for assigning relative weights to specific Virtual Reality Quality of Experience factors—usability (task completion and timely visibility of multi-user updates), 3D perception quality, motion-to-photon latency for view changes and actions, post-teleportation delay, headset energy consumption and temperature, and frame quality and its spatial/temporal variability—in order to aggregate them into a single quantitative QoE score for VR applications.

Background

Unlike stored 2D video streaming, VR Quality of Experience depends on a diverse set of factors that include interactive responsiveness, latency-sensitive immersion, and device constraints, in addition to traditional frame quality. The authors note that VR users act in complex and unpredictable ways and may teleport, making QoE assessment substantially more challenging than for video.

While some studies have attempted to infer weights via user studies, there is no accepted approach for combining the multiple VR-specific QoE dimensions into a single metric. Establishing such a weighting scheme is crucial for systematic design and evaluation of rendering, adaptation, prefetching, caching, and transport strategies in VR systems.