Visual closed-loop dynamics via ocular drift

The human visual system perceives its environment via eye movements, which are typically classified as saccades and drifts. Saccades are quick transitions of the gaze from one Region of Interest (ROI) to another and drifts are slower scanning motions in each ROI. Here we examine two contrasting schemes of perception via these ocular movements: an open-loop computational and a closed-loop dynamical scheme. The two schemes entail contrasting predictions, specifically regarding the kinematics of ocular drifts and their dependency on concurrent visual inputs. We have thus designed an experiment in which measuring ocular kinematics while modulating the available visual information (controlled by image size and by a gaze-contingent display) can discriminate between the two schemes. Our results reveal that vision is a closed-loop dynamic process. This is demonstrated by (i) a dependency of the drift trajectory on the concurrent visual input (ii) condition-specific convergence of the drift speed within