September 2019
Volume 19, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2019
Evidence for closed-loop visual acquisition
Author Affiliations & Notes
  • Liron Zipora Gruber
    Department of Neu-robiology, Weizmann Institute of Science, Rehovot 76100, Israel
  • Ehud Ahissar
    Department of Neu-robiology, Weizmann Institute of Science, Rehovot 76100, Israel
Journal of Vision September 2019, Vol.19, 146d. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Liron Zipora Gruber, Ehud Ahissar; Evidence for closed-loop visual acquisition. Journal of Vision 2019;19(10):146d.

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

The visual system acquires visual information via eye movements, which are typically classified as saccades (quick transitions of the gaze from one region of interest, ROI, to another) and drifts (slow scanning motions in each ROI). Two contrasting perceptual schemes can be consistent with this movement pattern: a computational and a dynamical scheme. In both schemes ROI selection obeys, at least in part, closed-loop dynamics in which target selection depends on the accumulated visual information. The computational scheme, however, assumes that at each ROI the visual system computes an internal representation of the image through an open-ended sequence of computations, much like in computer vision. The dynamical scheme, on the other hand, suggests that the entire process is inherently closed-loop and that there are no open-loop computations at any level. Instead, the system dynamically converges towards (though never reaches) perceptual steady-states in all its levels, including those controlling the ocular drift at each ROI. To test the predictions of the two schemes, we measured ocular behavior while modulating the available spatial information (by reducing image size and by mimicking tunnel vision). We show that the ocular drift (i) dynamically converges to a condition-specific speed, (ii) converges anew after each saccade within < 100 ms, (iii) exhibits increased periodic activity around 10 Hz when vision challenged and (iv) its trajectory depends on concurrent sensory data. This behavior strongly challenges the open-loop scheme and suggests that “early” visual processes are accomplished dynamically within low-level retinal-neuronal-ocular loops. Consistent with this suggestion, the visual system maintained selected acquisition strategies across conditions, via a coordinated control of saccade-rate and drift kinematics. Importantly, part of the saccade-drift coordination occurred at the resolution of individual saccades. These results suggest that vision is inherently a closed-loop process, which dynamically and continuously links all brain levels to their environment.


This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.