Although VEs are a useful tool for investigating human behaviors, there are important differences between VEs and the real world. Understanding and identifying these differences can help scientists better utilize VEs as an effective research tool while also helping to improve the technology and software. In immersive VEs often a person’s actions are limited or altered as a direct consequence of the interface. For instance, experiments have determined that even within an immersive HMD-VE people walk differently compared to the real world (Mohler). These results suggest that increasing the vertical field of view of HMDs and decreasing the weight of HMDs is an important factor if real-world walking behavior is desirable. In addition, in a VE people rarely see representations of themselves or other people due to the technical difficulty in rendering and accurately animating human bodies in real time. With our large tracking hall and in-house developed software we have been able to investigate the influence of having a virtual representation of oneself (self-avatar) on egocentric distance perception (Mohler). We found that when an observer is able to see a self-avatar they judge egocentric distances more accurately compared to the underestimation of distance observed when a self-avatar is absent. Further, it was also shown that the articulation of the self-avatar and not the visual information provided by the first-person perspective of the body was the critical factor that helped improve distance perception. These results suggest that there is a benefit to rendering a self-avatar in immersive VEs.

◘ Locomotion Behavior and Avatars in Immersive Virtual Environments

Locomotion Behavior and Avatars in Immersive Virtual Environments

It has been shown that people perform differently in Virtual Environments (VEs) as compared to the real world (at least for egocentric distance judgments). In addition, most immersive VEs currently contain little to no social content due to the complexity in rendering people and dynamic interactive behavior in real-time. In this research we work to increase the content of VE by adding virtual avatars, including by initially investigating the impact of a fully-articulated self-avatar on egocentric distance perception.

Our research aim is to improve the content and realism of immersive VE experiences by studying human perception and behavior.

Currently we have conducted two experiments using an egocentric distance judgment paradigm where individuals were asked to view a target location in virtual space and then close their eyes and walk to the previously seen target. In one case we used this paradigm to evaluate biomechanical differences between when people walk towards a seen or previously seen target in the virtual world as compared to the real world. Using a large tracking hall equipped with 16 Vicon Cameras we found that when wearing a head-mounted display (HMD) people walk more slowly, with a shorter stride length and a wider step width [1]. In the second experiment we evaluated the impact of having a fully-articulated self-avatar on subsequent egocentric distance judgments. We found that viewing an articulated self-avatar as compared to a line on the floor indicating your location in space improved egocentric distance judgments within a HMD-VE (see Figures 1 and 2) [2].

Figure 1: From left: viewer’s hands (1) and feet (2) in virtual world,
and virtual world where the target and hallway were viewed before
blind-walking (3).

Figure 2: A participant wearing reflective markers, a backpack and a HMD
in our tracking hall (left), the ViconIQ Skeleton which indicates the
person’s joint and body locations in the tracking hall in real-time
(middle) and the visual representation of this person (avatar) within
the virtual world (right).

For future work we plan to further investigate the reasons for a self-avatar influencing egocentric distance judgments in a VE. Specifically we will investigate the impact of near versus far space and the articulation of the avatar. In addition we will develop a library of animations/avatars which can be used to dynamically add animated avatars into the virtual world. Adding this social content to a VE should allow one to explore many research topics including, but not limited to social behavior, embodied cognition and adaptation of specific motor and social behavior to modified visual information about one’s movements and appearance.

REFERENCE
◘ B. Mohler, J. Campos, M. Weyel and H. H. Bülthoff: Gait parameters while walking in a head-mounted display virtual environment and the real world., Proceedings of the 13th Eurographics Symposium on Virtual Environments and 10th Immersive Projection Technology Workshop (IPT-EGVE 2007), 85-88 (07 2007).

◘ B. Mohler, H. H. Bülthoff, W. B. Thompson and S. H. Creem-Regehr: A full-body avatar improves egocentric distance judgments in an immersive virtual environment. Proceedings of the 5th Symposium on Applied Perception in Graphics and Visualization (APGV 08), 194-197. (Eds.) Spencer, S. N. ACM Press, New York, NY, USA (08 2008)