GestureDrawer

Abstract

Existing 1 empty-handed mid-air interaction techniques for system control are typically limited to a confined gesture set or point-and-select on graphical user interfaces. In this paper, we introduce GestureDrawer, a one-handed interaction with a 3D imaginary interface. Our approach allows users to self-define an imaginary interface, acquire visuospatial memory of the position of its controls in empty space and enables them to select or manipulate those controls by moving their hand in all three dimensions. We evaluate our approach with three user studies and demonstrate that users can indeed position imaginary controls in 3D empty space and select them with an accuracy of 93% without receiving any feedback and without fixed landmarks (e.g. second hand). Further, we show that imaginary interaction is generally faster than mid-air interaction with graphical user interfaces, and that users can retrieve the position of their imaginary controls even after a proprioception disturbance. We condense our findings into several design recommendations and present automotive applications. CCS CONCEPTS • Human-centered computing~Gestural input • Humancentered computing~User interface design • Human-centered computing~User centered design KEYWORDS

Eliciting Contact-Based and Contactless Gestures With Radar-Based Sensors

“…GestureDrawer [12] introduces a novel one-handed interaction technique that enables users to place imaginary user interface controls in 3D empty space without receiving any kind of feedback (completely eye-free and device-free) and without fixed landmarks.…”

Pocket6

“…Similarly to [4], participants faced two squares displayed on the screen. These squares had to be aligned by a 3D drag-and-drop gesture.…”