Fishbowl Fitts’ Law
In this paper, Arsenault and Ware report on the effects of head tracking and stereo imaging for eye-hand coordination in a mixed reality environment.
The authors use a Fitts’ law type experiment to study the effect on reaction times when subjects are forced to move their heads to look around a barrier in an Fishtank VR setup. Would head tracking alone, stereo imaging, or an interaction of the two have any significance in the time it takes for a person to accurately move a target armature?
Fishtank is a VR technique able to provide fine angular detail in a small contained work area. There is a mirror placed in the field of view of the operator which blocks their manual work area. This mirror reflects the image of a high resolution monitor so that the view appears to show virtual objects below the glass panel in the workspace below. An armature called PHANToM provides a tangible interaction point. It is tracked mechanically and rendered in the view as a 3D virtual cursor. The user looks down and through the mirror to the interaction area and is asked to perform target touching tasks. A headworn device is used to track the users point of view, and has alternating shutters to provide stereoscope view by coordinating the rate and phase of the mirrored monitor image with the eye shutters.
With this setup, the investigators had control of several of the important cues we use to make sense of eye-hand interaction spaces. These important cues are usually cited as: stereoscopic depth, motion parallax. occlusion, and perspective. There was the ability to simulate motion parallax (differential movement of objects relative to the position of the viewer) through head tracking, and the ability to render the 3D space stereoscopically and with single POV to simulate perspective effects. Occlusion problems were not investigated since the mirror completely obscured the real work area.
They found that there is significant benefit in 3D tasks to both stereo and eye-coupled perspective, but that stereo was much more responsible for improvement in performance and in adjustment to the experiment setup. It adds convincing evidence for the hypothesis that humans can adjust very quickly to odd and distorted perspectives and that stereoscopic vision plays a larger role than motion parallax in performing these adjustments.
This paper is important because it very confidently demonstrates the observed main effects of both head tracking and stereoscopic vision on eye-hand movement acuity. By extending and replicating Fitts’ Law it has very high external validity.
Comments are closed.