It is often assumed that more realism is always desirable. In particular, many techniques for locomotion in Virtual Reality (VR) attempt to approximate real-world walking. However, it is not yet fully understood how the design of more realistic locomotion techniques influences effectiveness and user experience. In the previous VR studies, the effects of interaction fidelity have been coarse-grained, considering interaction fidelity as a single construct. We argue that interaction fidelity consists of various independent components, and each component can have a different effect on the effectiveness of the interface. Moreover, the designer's intent can influence the effectiveness of an interface and needs to be considered in the design. Semi-natural locomotion interfaces can be difficult to use at first, due to a lack of interaction fidelity, and effective training would help users understand the forces they were feeling and better control their movements. Another method to improve locomotion interaction is to develop a more effective interface or improve the existing techniques. A detailed taxonomy of walking-based locomotion techniques would be beneficial to better understand, analyze, and design walking techniques for VR.
We conducted four user studies and performed a meta-analysis on the literature to have a more in-depth understanding of the effects of interaction fidelity on effectiveness. We found that for the measures dependent on proprioceptive sensory information, such as orientation estimation, cognitive load, and sense of presence, the level of effectiveness increases with increasing levels of interaction fidelity. Other measures which depend more on the ease of learning and ease of use, such as completion time, movement accuracy, and subjective evaluation, form a u-shape uncanny valley. For such measures, moderate-fidelity interfaces are often outperformed by low- and high-fidelity interfaces.
In our third user study, we further investigated the effects of components of interaction fidelity, biomechanics and transfer function, as well as designers' intent. We learned that the biomechanics of walking are more sensitive to changes and that the effects of these changes were mostly negative for hyper-natural techniques. Changes in the transfer function component were easier for the user to learn and to adapt to. Suitable transfer functions were able to improve some locomotion features but at the cost of accuracy.
To improve the level of effectiveness in moderate-fidelity locomotion interfaces we employed an effective training method. We learned that providing a visual cue during the acclimation phase can help users better understand their walking in moderate-fidelity interfaces and improve their effectiveness. To develop a design space and classification of locomotion techniques, we designed a taxonomy for walking- based locomotion techniques. With this taxonomy, we extract and discuss various characteristics of locomotion interaction. Researchers can create novel locomotion techniques by making choices from the components of this taxonomy, they can analyze and improve existing techniques, or perform experiments to evaluate locomotion techniques in detail using the presented organization. As an example of using this taxonomy, we developed a novel locomotion interface by choosing a new combination of characteristics from the taxonomy. / Ph. D. / Virtual Reality researchers have been trying to develop natural travel techniques to allow users to physically walk and move in virtual environments rather than using unnatural methods such as joysticks. Using such techniques, the user can physically move or perform actions similar to walking to navigate through virtual environments. More natural travel methods can improve various parameters such as sense of presence and spatial understanding. However, the effects of increasing naturalness of walking on the user experience have not been known for years.
In this dissertation, we have run four user experiments, performed a meta-analysis on the literature and developed a taxonomy to contribute a better understanding of how increasing levels of walking naturalness can affect user experience in Virtual Reality. Our findings can benefit designers and researchers in designing novel travel techniques, improve existing techniques or more in-depth understating of what to expect when employing a certain travel technique.
Identifer | oai:union.ndltd.org:VTETD/oai:vtechworks.lib.vt.edu:10919/74945 |
Date | 06 February 2017 |
Creators | Nabiyouni, Mahdi |
Contributors | Computer Science, Bowman, Douglas A., North, Christopher L., Polys, Nicholas F., Gracanin, Denis, Hollerer, Tobias |
Publisher | Virginia Tech |
Source Sets | Virginia Tech Theses and Dissertation |
Detected Language | English |
Type | Dissertation |
Format | ETD, application/pdf, application/pdf |
Rights | In Copyright, http://rightsstatements.org/vocab/InC/1.0/ |
Page generated in 0.0021 seconds