Abstract
Virtual reality applications depend on multiple factors, for example, quality of rendering, responsiveness, and interfaces. In order to evaluate the relative contributions of different factors to quality of experience, post-exposure questionnaires are typically used. Questionnaires are problematic as the questions can frame how participants think about their experience and cannot easily take account of non-additivity among the various factors. Traditional experimental design can incorporate non-additivity but with a large factorial design table beyond two factors. Here, we extend a previous method by introducing a reinforcement learning (RL) agent that proposes possible changes to factor levels during the exposure and requires the participant to either accept these or not. Eventually, the RL converges on a policy where no further proposed changes are accepted. An experiment was carried out with 20 participants where four binary factors were considered. A consistent configuration of factors emerged where participants preferred to use a teleportation technique for navigation (compared to walking-in-place), a full-body representation (rather than hands only), the responsiveness of virtual human characters (compared to being ignored) and realistic compared to cartoon rendering. We propose this new method to evaluate participant choices and discuss various extensions.
Highlights
The construction of virtual reality (VR) applications is a complex task involving the choice of appropriate hardware and royalsocietypublishing.org/journal/rsos R
VR is rapidly moving towards becoming a major consumer product, with high-quality head-mounted display (HMD) systems available at costs cheaper than many smartphones and with millions of sales throughout the world
Application designers face a number of trade-offs—for example, should the environment look as photorealistic as possible, or be more cartoon-like? The virtual space in which people can move will typically be much larger than the physical space in which they operate, so an interface method is required to enable movement around the virtual space
Summary
The construction of virtual reality (VR) applications is a complex task involving the choice of appropriate hardware and royalsocietypublishing.org/journal/rsos R. In [27], we exploited this analogy by introducing a methodology for the measurement of presence based on colour matching theory, which did not require questionnaires at all In this method, participants were introduced to a number of factors in the VR that they could modify during the course of their exposure—field of view, whether or not they had a virtual body that moved synchronously with their own movements, whether they saw the virtual environment from a first- or third-person perspective, and quality of visual rendering. The second, and most important difference, is that proposals for change of factor levels are chosen by the RL, which introduces a clear stopping rule—when no more changes are accepted by the participant—indicating convergence of the RL, in the sense that additional changes would not improve participant preference We demonstrate this method in an experiment involving four binary factors. After a period of training where participants learned about these four factors, they made their way to the concert, and every so often the RL proposed a binary choice to either change the level of one of the four factors or leave the current configuration unchanged
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
