A Comparison of Virtual Reality Interaction Techniques for Retrieving Out-of-Range Objects and Optimization of a Tether-Handle Object Retrieval Technique

Abstract

Many interactive VR applications provide the ability to select, grab, and manipulate objects in the virtual environment, whether it be with a motion-tracked handheld controller or a gestural input system powered by hand-tracking. However, when objects leave the user's reach, standard grabbing interactions become less useful. In these cases, a distance selection and grabbing technique is needed. Most existing distance grabbing techniques use separate interaction metaphors from regular grabbing, such as pointing or hand-extension. However, to avoid introducing a brand new metaphor of object interaction used only in exceptional distance grabbing cases, and to avoid cluttering the controller or gestural interface with multiple functions, we created a distance selection and grabbing technique that leverages the interaction techniques users already use to manipulate nearby objects. Our technique attaches a grabbable handle tethered to each distant object.

We then ran a two-phase study with 52 subjects. In the first phase, we compared four distance retrieval techniques, including our novel technique and our own implementations of existing techniques. We had subjects complete a series of game-like tasks in VR, using each of the techniques, and we collected the subjects' opinions and performance data. In the second phase, we compared similar variations of our technique, each variation making some design trade-off or specializing the technique for a specific use-case.

The results of our analysis suggest that, overall, our tether-handle technique is preferred to hand-extension and on-par with the other techniques we evaluated. With hand-tracking, however, our technique performed slightly worse and was preferred slightly less than the other techniques. Based on the results from our second phase of the study, we've determined that variations on our technique that reduced visual clutter were strongly preferred. We propose a follow-up that would repeat the technique comparison, incorporating relevant changes to optimize our technique.