Gesture select:: acquiring remote targets on large displays without pointing

When working at a large wall display, even if partially utilized, many targets are likely to be distant from the user, requiring walking, which is slow, and interrupts workflow. We propose a novel technique for selecting remote targets called Gesture Select, in which users draw an initial mark, in a target's direction; rectilinear gestures represented as icons are dynamically overlaid on targets within a region of interest; the user then continues by drawing the continuation mark corresponding to the target, to select it. Extensions to this technique to support working with remote content for an extended period, and learning gesture shortcuts are presented. A formal experiment indicates Gesture Select significantly outperformed direct selection for mid/far targets. Further analysis suggests Gesture Select performance is principally affected by the extent to which users can read the gestures, influenced by distance and perspective warping, and the gesture complexity in the ROI. The results of a second 2-D experiment with labeled targets indicate Gesture Select significantly outperformed direct selection and an existing technique.