Designing a Gaze Gesture Guiding System
We propose the concept of a guiding system specifically designed for semaphoric gaze gestures, i.e. gestures defining a vocabulary to trigger commands via the gaze modality. Our design exploration considers fundamental gaze gesture phases: Exploration, Guidance, and Return. A first experiment reveals that Guidance with dynamic elements moving along 2D paths is efficient and resistant to visual complexity. A second experiment reveals that a Rapid Serial Visual Presentation of command names during Exploration allows for more than 30% faster command retrievals than a standard visual search. To resume the task where the guide was triggered, labels moving from the outward extremity of 2D paths toward the guide center leads to efficient and accurate origin retrieval during the Return phase. We evaluate our resulting Gaze Gesture Guiding system, G3, for interacting with distant objects in an office environment using a head-mounted display. Users report positively on their experience with both semaphoric gaze gestures and G3.
William Delamare, Teng Han, and Pourang Irani. 2017. Designing a gaze gesture guiding system. In Proceedings of the 19th International Conference on Human-Computer Interaction with Mobile Devices and Services (MobileHCI '17). ACM, New York, NY, USA, Article 26, 13 pages. https://doi-org.uml.idm.oclc.org/10.1145/3098279.3098561