Virtual reality (VR) technology has seen remarkable development and resurgence over the last few years, which is reflected in both enterprise applications and academic research. Chief among this new wave of VR technology is so-called room-scale VR, wherein sensors allow a user to navigate a calibrated space inside of a virtual world using their natural body movement. Since these sensors know how tall a user is, users retain their physical height inside of room-scale software applications. Suddenly, software development has to account for varying user heights. This is particularly interesting in multi-user scenarios, where interpersonal height differences from the physical world now affect interpersonal relationships in the virtual world. However, this novel feature of the technology also grants opportunities to software developers and designers, who gain the ability to treat user height and scale as a design variable and alter it for the needs of their application.
In this thesis, we seek to contribute towards exploring this largely unexplored phenomenon by determining how to design and implement room-scale VR software that uncovers the effects of user scaling in multi-user room-scale VR scenarios. Following a literature review of the latest proceedings from major academic VR conferences, we set out to design a study towards this purpose. To achieve that, we build a theoretical understanding of VR, which we then shape into an interaction model to help us design the study. Since our study design requires a custom multi-user room-scale VR application, we both design and implement this network application, using the advanced 3D graphics engine, Unity, for our software development.
The experimental design of the study involves a dyadic relationship wherein two participants are exposed to a series of conditions in room-scale VR. The first two conditions serve to shape a common understanding of typical interaction techniques found in room-scale VR applications, to ensure that users have a common frame of reference regardless of previous experience with the technology. Then, we expose one user group to an experimental condition in which one user retains their regular size in relation to the environment, while one user is scaled down and becomes tiny; the large user has to aid the small user in navigating a maze. Another group of participants serves as the control group for the effects of the scaling, and is only exposed to the first two conditions.
During development, we test various aspects of our experimental design in two testing sessions involving N=18 and N=23 users. We also include a small, preliminary pilot study based on our final study design, producing one set of study data for each user group.
In conclusion, we successfully solve how to design and implement room-scale VR software that uncovers the effects of user scaling in multi-user room-scale VR scenarios. Moreover, while our pilot study results are preliminary in nature, they are promising towards indicating that altering the scale of users impacts the way dyadic relationships are experienced inside of room-scale VR applications.
|Uddannelser||Datalogi, (Bachelor/kandidatuddannelse) Kandidat|
|Udgivelsesdato||3 apr. 2018|
|Vejledere||Magnus Rotvit Perlt Hansen & Mads Rosendahl|
- Virtual Reality
- Computer Science
- Room-scale VR
- HTC Vive
- Interaction model
- User study
- Likert scale