To write not select, a new text entry method using joystick

Abstract

Existing joystick text entry methods for game and TV boxes are curser-based selections on virtual keyboards. In this paper we present a new text entry method using joysticks as tangible devices to capture users’ freehand writing gestures. The method has considerable accuracy to accomplish English text entry. On the prediction model, we introduced HMM algorithm so users can enter text assisted with automatic correcting. We conducted a pairwise usability test on the keyboard selection method and writing-with-joystick method. The result shows that both of them are very easier to learn and writing-with-joystick is faster than the keyboard selection method both on the prediction model or none-prediction model. Subjects also report that using the keyboard selection method to enter text can be boring when using handwriting is somehow natural. This result indicates that writing with joystick may be another text entry option for game console or Smart TV users.