Visual Perception in Simulated Reality

Abstract

This thesis addresses the use of features of human visual perception to alleviate computation time for real-time computer graphics applications and the initial steps towards the construction of a perceptually informed virtual reality environment. Using a subset of gaze-contingent rendering techniques, named perceptually lossless foveated rendering techniques, real-time rendering systems are able to selectively render at much lower fidelity in a way that the average user is unable to distinguish any difference in quality. This is achieved through the use of an eye tracking device, in order to render a fixed region around the user's point of gaze on a display surface at high quality with the rest of the render (which is in their peripheral field of view) at lower spatial quality.Although foveated rendering techniques have been explored in the past, it is only more recently that eye tracking and supporting rendering hardware have reached a point which reliably enables development of perceptually lossless (indistinguishable) investigation. Additionally, the resurgence of virtual reality in the commercial sector (along with its demands for rendering quality that exceeds the capability of many modern, commercially-accessible hardware) justifies the study and adoption of these methods.As such, this thesis presents the work conducted as part of an Engineering Doctorate in Digital Media with the Centre for Digital Entertainment in collaboration with Disney Research on the construction of a commercially accessible, enthusiast level, perceptually augmented virtual reality environment. This includes research towards novel perceptually lossless foveated rendering methods, addressing concerns specific to gaze-contingent methods, some of the engineering efforts towards the construction of the virtual reality environment, and additional related work conducted specifically as part of the collaboration with Disney Research.

Date of Award	23 Nov 2016
Original language	English
Awarding Institution	University of Bath
Supervisor	Darren Cosker (Supervisor) & Kenny Mitchell (Supervisor)