Path integration from optic flow and the role of eye movements

Abstract

Abstract Optic flow provides critical visual information for self-motion perception, including direction (heading), speed, and travel distance (path integration). While eye movements are naturally elicited during optic flow, most lab-based studies of human path integration require central fixation, potentially altering perception. Here, we examined whether free eye movements improve distance discrimination compared with fixation, testing the hypothesis that oculomotor signals may support oculomotor odometry . The participants viewed two consecutive optic flow stimuli simulating forward self-motion across a ground plane and judged which interval represented the longer travel distance. Eye movement instructions varied across the four conditions, allowing either free-viewing or requiring fixation (still allowing for miniature slow and fast eye movements within a control window) during each interval. Behavioral accuracy was analyzed with generalized linear mixed-effects models, and eye movements were quantified either including or excluding saccades. As expected, free-viewing induced larger and faster saccades and higher smooth eye-movement speeds than fixation did. However, contrary to our hypothesis, fixation did not systematically impair distance discrimination. While cumulative eye-movement amplitudes scaled with travel distance as predicted, psychometric functions for perceived distance were similar across conditions. These findings, together with results from other recent studies, suggest that fixation-based paradigms do not impair the study of certain aspects of self-motion perception, including settings requiring stable gaze (e.g., EEG recordings).