Human-Centered Fidelity Metrics for Virtual Environment Simulations

It is increasingly important to provide fidelity mecrics for rendered images and interactive virtual environments (VEs) targeting transfer of training in real-world task situations. Computational metrics which aim to predict the degree of fidelity of a rendered image can be based on psychophysical observations. For interactive simulations, psychophysical investigations can be carried out into the degree of similarity between the original and a synthetic simulation. Psychophysics comprises a collection of methods used to conduct non-invasive experiments on humans, the purpose of which is to study mappings between events in an environment and levels of sensory responses to those events. This tutorial will present the techniques and principles towards conducting psychophysical studies, for assessing image quality as well as fidelity of a VE simulation and how results from such studies contribute to VE system design as well as to computational image quality metrics. Methods based on experiments for measuring the perceptual equivalence between a real scene and a computer simulation of the same scene will be reported. These methods are presented through the study of human vision and include using photorealistic computer graphics to depict complex environments and works of art. In addition, physical and psychophysical fidelity issues in the assessment of virtual environments will be emphasised. Specifications for correct matching between the psychophysical characteristics of the displays and the human users’ sensory and motor systems will he discussed as well as some examples of the consequences when systems fail to be physically well matched to their users. Complete experimental cycles will be described from the initial idea and design, to pilot study, experimental redesign, data collection, analysis and post-experiment lessons learned. Examples will include research on spatial orientation in Virtual Reality, assessing fidelity of flight simulators, fidelity of simulation of humans and clothing and measuring perceptual sensitivity to latency. This tutorial requires no fundamental prerequisites. It would help if the attendee had some knowledge of experimental design, and of some personal experience of computer graphics and simulation systems. However, the course will be self-contained