Dimensions of perceptual-driven rendering in Virtual Reality

EPSRC research area: Graphics and visualisation

The recent availability of high-performance graphics technology and consumer-grade head-mounted-displays has meant that Immersive Virtual Reality (VR) is becoming increasingly popular both for research, leisure and commercial purposes. Generating VR experiences is notoriously computationally expensive, with demand and expectations always outstripping capability.

The high level aim of this project is to investigate the features of the human perceptual and cognitive systems that can be exploited to reduce the computational cost on various components of a VR system in order to improve the quality of the virtual experience for given hardware. One example of this would be lowering the rendering quality as a function of eccentricity away from the fovea (focus point on the human eye retina).
The main objectives are to:
Establish a reference architecture for the VR pipeline
Define a taxonomy that relates features of perception/cognition to components of the reference architecture
Perform a systematic review of the current state-of-art in terms of perceptual and cognition driven rendering for VR, identifying areas of previous success or failure and areas for promising exploration
Conduct experiments in these areas to determine whether meaningful savings/improvements are possible

Assuming areas can be identified that suggest meaningful savings/improvements, components of the VR pipeline will be re-engineered to embody these principles, and further experiments conducted to confirm results.