Gestural Workspaces for Computer Interaction

Abstract

While the hardware challenges facing integration of hand-gestural controls into mainstream computer interfaces appear to be shrinking, design challenges related to the physical and mental burden required of gesture interfaces remain. This study aims to determine the effect of the gestural workspace configuration (vertical, horizontal, and jointspace) and affordance of arm support on speed, accuracy, and performance when performing pointing tasks on a computer. Seventeen participants played a computer card game (Solitaire) and then completed Fitts’ serial clicking tasks to evaluate performance in five gestural workspace configurations implemented using a Microsoft Kinect®. Gestural input configurations were also compared to a traditional mouse input. A traditional mouse performed better than gestural controls in all aspects of performance. Among gestural workspaces, the vertical configuration performed significantly better in throughput and accuracy outcomes. Limitations of the gestural tracking system and large variation in individual performance may hinder establishment of generalized design recommendations at this point; however, gestural workspace configurations with direct mapping to onscreen movements and the presence of supporting surfaces appear to increase user performance and reduce perceived difficulty.