Gaze Plus Gesture Interface: Considering Social Acceptability

In public places like cafes, the usage of smart glasses with interfaces including controller, touchpad, voice, or mid-air gesture not only receives a lot of interest from people around him or her but also cannot protect the privacy of individuals. If smart glasses become more advanced and more popular than regular glasses in the future, socially acceptable user interfaces can be required. In this paper, we propose a user interface on HoloLens by using gaze tracking instead of head tracking for navigation, and unobtrusive gesture based on deep learning instead of mid-air gestures for selection/manipulation, that is more socially acceptable than the existing user interfaces on smart glasses. A study was conducted to investigate social acceptability from the users' perspective, and the results showed the advantages of the proposed method to improve social acceptability.