Cyclopean, Dominant, and Non-dominant Gaze Tracking for Smooth Pursuit Gaze Interaction.

User-centered design questions in gaze interfaces have been explored in multitude empirical investigations. Interestingly, the question of what eye should be the input device has never been studied. We compared tracking accuracy between the "cyclopean" (i.e., midpoint between eyes) dominant and non-dominant eye. In two experiments, participants performed tracking tasks. In Experiment 1, participants did not use a crosshair. Results showed that mean distance from target was smaller with cyclopean than with dominant or non-dominant eyes. In Experiment 2, participants controlled a crosshair with their cyclopean, dominant and non-dominant eye intermittently and had to align the crosshair with the target. Overall tracking accuracy was highest with cyclopean eye, yet similar between cyclopean and dominant eye in the second half of the experiment. From a theoretical viewpoint, our findings correspond with the cyclopean eye theory of egocentric direction and provide indication for eye dominance, in accordance with the hemispheric laterality approach. From a practical viewpoint, we show that what eye to use as input should be a design consideration in gaze interfaces.