Analysis and comparison of target assistance techniques for relative ray-cast pointing

Pointing at displays from a distance is becoming a common method of interacting with computer applications and entertainment systems, using devices such as the Wii Remote, the PlayStation Move controller, or the Microsoft Kinect. These systems often implement relative forms of ray-cast pointing, in which the user simply points a hand-held input device towards targets on the screen. Ray-casting interaction is easy for novices to learn and understand, but this technique often suffers from accuracy problems: for example, hand jitter, arm fatigue, calibration drift, or lack of skill can all reduce people's ability to acquire and select on-screen targets. In this paper, we analyse and evaluate the idea of target assistance as a way to address the accuracy problems of ray-cast pointing. Although several assistance schemes have been proposed for mouse-based pointing, these ideas have not been tested in distant-pointing settings, and there is little knowledge available to guide design in this increasingly common interaction scenario. To establish this basic design knowledge, we carried out four studies of relative ray-casting using three different target assistance techniques-two motor-space techniques (sticky targets and a novel form of target gravity), and one acquisition-feedback technique that combined visual, tactile, and auditory feedback. Our first three studies tested each assistance technique separately, to explore how different parameters for each method affected performance and perceptibility. Our fourth study carried out a direct comparison of the best versions of each technique, and also examined the effects of distractor objects placed in the path to the target. Our studies found that the two motor-space techniques were extremely effective in improving selection accuracy without being highly obvious to users, and that the new gravity-based technique (which attracts the cursor even when it is not over the target) performed best of all. There was no observed effect on performance when the combined acquisition-feedback technique was used. Our studies are the first to comprehensively explore the optimization, performance, and perceptibility of target assistance techniques for relative ray-casting-our results provide designers with clear guidelines about what methods to use, how to configure the techniques, and what effects can be expected from their use. (C) 2013 Elsevier Ltd. All rights reserved.