August 2012
Volume 12, Issue 9
Vision Sciences Society Annual Meeting Abstract  |   August 2012
Active visual sampling in uncertain environments
Author Affiliations
  • David Evens
    School of Experimental Psychology, University of Bristol
  • Tom Cassey
    Department of Computer Science, University of Bristol
  • James Marshall
    Department of Computer Science, University of Bristol
  • Rafal Bogacz
    Department of Computer Science, University of Bristol
  • Casimir Ludwig
    School of Experimental Psychology, University of Bristol
Journal of Vision August 2012, Vol.12, 618. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      David Evens, Tom Cassey, James Marshall, Rafal Bogacz, Casimir Ludwig; Active visual sampling in uncertain environments. Journal of Vision 2012;12(9):618.

      Download citation file:

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

  • Supplements

Eye movements serve to gather information from the environment. Typically, stimuli cannot be sampled in parallel, meaning we only gain knowledge about the currently sampled source of information. We examined whether, given limited time to inspect a number of sources of noisy information, the time allocated to each source is adapted to the level of uncertainty in both the environment and sensory signals. Observers performed a relative motion direction discrimination judgement between two random dot kinematograms (RDKs). Motion information was only delivered when gaze was fixed on a pattern. Low and high coherence patterns were paired independently, but when the coherence of the two patterns differed, the more noisy source should be inspected for longer. In one condition, the mapping between dot polarity and coherence was random: when inspecting the first pattern, the noise level in the second pattern was unknown. In a second condition, the coherence was consistently mapped onto the polarity of the RDK dots and observers have perfectly reliable information about the sensory noise they will encounter. In this case, we expected more complete adaptation of sampling time to the noise levels, matched to the precision of the underlying internal direction estimates. Observers sampled the patterns adaptively, but there was no difference in the extent of adaption under random and consistent mappings: observers ignored the available information about the noise in the information sources. In a second experiment, we provided this information further in advance, but sampling behaviour remained unaffected. The degree of adaptation was smaller than expected based on independent measurements of the precision in the internal sensory direction estimates. We conclude that observers based their gaze allocation solely on the uncertainty of the first source they inspected. The sampling strategy was not adjusted to the global level of (un)certainty about the environment.

Meeting abstract presented at VSS 2012


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.