Gestures for Controlling a Moveable TV

We investigate the effects of physical context on the preference and production of touchless (3D) gestures, focusing on what users consider to be natural and intuitive. Using an elicitation task, we asked for users' preferred gestures to control a "moving TV" display from a distance of 3-4m. We conducted three user studies (N=16 each) using the same premise but varying the physical conditions encountered, such as number of hands available or distance and orientation to the display. This is important to ensure the robustness of the gesture set. We observed two dominant strategies which we interpret as dependent on the user's mental model: hand-as-display and hand-moving-display. Across the varying conditions, users were found to be consistent with their preferred gesture strategy while varying the production (number of hands, orientation, extension of arms) of their gestures in order to match both their mental models and the physical context of use. From a technology perspective, this natural variation challenges the notion of identifying "the optimal gesture set" and should be taken into account when designing future systems with gesture control.