Adaptive Generation of Emotional Impact Using Enhanced Virtual Environments

Abstract

In this paper we introduce novel methods of intensifying and varying the user experience in virtual environments (VE). VEs technically have numerous means for crafting the user experience. Little has yet been done to evaluate those means of expression (MoEs) for their emotional impact on people and to use their capability to create different experiences and subtly guide the user. One of the reasons is that this requires a system which is capable of easily and dynamically providing those MoEs in such a way that they can easily be composed, evaluated, and compared between applications and users. In the following, we first introduce our model of both informational and emotional impact of VEs on users, introduce our dynamic, expressive VR-system, and present our novel evaluation and rating method for MoEs. MoEs can be used to guide attention to specific objects or build up an emotion or mood over time. We then present a study in which users experience 30 selected MoEs and rate their qualitative emotional impact using this rating method. We found that different MoEs can be used to elicit many diverse emotions which were surprisingly consistent among the test persons. With these results, our work enables new ways to make VEs more interesting and emotionally engaging, especially over a longer period of time, opening new possibilities, for example, to increase the motivation for long, stressful, and tiresome training as in neurorehabilitation.