Improving Accuracy and Efficiency of Hand Gestures on Multi-Touch Devices

Geometric transformation gestures such as rotation, scaling and dragging are extremely common. There are multiple variants of designing and implementing these gestures. Variants include slightly modifying the gesture input (e.g. different original placement or tracing of fingers) or the resulting action (e.g. scale factor, retention of scale centre or rotation degree). There has not been a significant amount of research assessing the best design of geometric transformation gestures across multiple multi-touch devices. We describe our research project that looks at variants of standard geometric transformation hand gestures. We hypothesise that these variants are superior to standard geometric transformation gestures (in terms of supporting more precise transformations and faster completion times) and are as easy to initiate and maintain as the standard gestures. We also discuss our experiences in implementing these variants and describe and present user experiments we have completed in order to test our hypotheses. The results show that only some of our variants are more precise and support faster transformation completion and that only some of these results are mirrored between devices. Furthermore, only some of our variants are as easy to initiate and maintain as the standard gestures.