A Comparison of Smooth Pursuit- and Dwell-based Selection at Multiple Levels of Spatial Accuracy
Dillon James Lohr, Oleg V. Komogortsev · 2017
In this paper, we present a smooth pursuit--based alternative to dwell-based selection for eye-guided user interfaces. Participants attempt to perform both dwell- and pursuit-based selections while we artificially reduce the spatial accuracy of an affordable eye tracker to see how resilient both selection methods are. We find that the time to perform a pursuit-based selection remains consistent even as spatial accuracy degrades, unlike dwell-based selection which takes considerably longer to perform the worse the spatial accuracy becomes. We argue that smooth pursuit--based selection will be important in eye-tracking systems with low spatial accuracy, such as very low cost trackers, certain self-made systems, and calibration-free systems.