Around-device finger input on commodity smartwatches with learning guidance through discoverability
Marium-E-Jannat, Xing-Dong Yang, Khalad Hasan · International Journal of Human-Computer Studies · 2023
User interactions with smartwatches are limited due to one-finger usage on small touch screens. However, smartwatch interaction capabilities can be extended beyond the screen into the around-device space by leveraging multi-finger interactions. In this paper, we explore suitable finger postures and gestures in mid-air and on the back of the palm to extend interaction capabilities on the unmodified commodity smartwatch while ensuring their learnability through discoverability. We conduct a design study to find a set of finger postures and gestures suitable for interacting with off-the-shelf smartwatches. An application is then designed to detect the posture-gesture set with on-device dual cameras while examining their classification accuracy. To facilitate learnability through discoverability of the gesture-posture set, we explore ‘context-aware-hints’ that shows a sub-set of gesture-posture and their associated actions based on the usage context. Results from a user study show that the guidance helps users to discover and learn them quickly and accurately.