Abstract
Previous research shows that when participants can use direct walking to control their movement through a small, HMD-based immersive virtual environment (IVE), they report a greater sense of ‘presence’ in that environment than when they must use a metaphoric or indirect action, such as stepping-in-place or pressing-a-button on a hand-held wand for locomotion control. However, when the IVE is larger than the available physical space, it becomes necessary to re-consider the use of alternative metaphors.
We introduce a novel method for naturalistic, augmented direct locomotion through large-scale IVEs: seven-league-boots, describing its technical implementation and discussing alternative options and parameters. We then present the results of a series of experiments that seek to provide qualitative and quantitative insight into the relative strengths and weaknesses of this method in comparison to three commonly-used alternatives: virtual flying/gliding, via a button-press on a wand; uniformly accelerated real walking, achieved by allowing the user to walk normally but applying a uniform gain to the output of the tracker that re-defines his corresponding position in the virtual world; and normal walking without gain, but with intermittent major adjustments of the location and orientation of the IVE relative to the real-world position of the participant.
Seven-league-boots locomotion is characterized by an exaggeration of -only- the component of a person's movement that is aligned with his direction of intended travel. This requires knowing when purposeful travel is intended, and accurately predicting its direction.
A within-subjects experiment with 8 naïve participants found significantly higher 7-point ratings of overall preference for IVE exploration via boots (µ=6.25, σ=0.71) than with uniform gain (µ=3.38, σ=1.92), flying (µ=4.25, σ=0.89), or interrupted walking (µ=3.88, σ=1.46). A comparison of the accuracy of self-localization with respect to occluded landmarks in IVEs explored via boots vs flying is forthcoming and will also be reported.
This research was supported by the National Science Foundation (IIS-0313226), by the University of Minnesota through a Digital Technology Center seed grant and by the Linda and Ted Johnson Digital Design Consortium Endowment and Lab Setup Funds.