July 2013
Volume 13, Issue 9
Vision Sciences Society Annual Meeting Abstract  |   July 2013
Subitizing and estimation emerge from a computational saliency map model of object individuation
Author Affiliations
  • Rakesh Sengupta
    Center for Neural and Cognitive Sciences, University of Hyderabad
  • S. Bapiraju
    Center for Neural and Cognitive Sciences, University of Hyderabad
  • David P. Melcher
    Center for Mind/Brain Sciences (CiMec), University of Trento
Journal of Vision July 2013, Vol.13, 235. doi:https://doi.org/10.1167/13.9.235
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Rakesh Sengupta, S. Bapiraju, David P. Melcher; Subitizing and estimation emerge from a computational saliency map model of object individuation. Journal of Vision 2013;13(9):235. doi: https://doi.org/10.1167/13.9.235.

      Download citation file:

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

  • Supplements

One of the principal tasks of vision is to extract information about individual objects and groups of objects from the retinal image. This has led to the suggestion that our core sense of number emerges from these basic perceptual processes (Feigenson et al., 2004, TRENDS in Cognitive Sciences, 8, 307-314). We extended a recent model of saliency maps in human parietal cortex (Roggerman et al, 2010, NeuroImage, 52, 1005-1014) to account for human numerosity judgements within both the subitizing and estimation ranges using a single parameter: the inhibition between connected neural nodes in the lateral intraparietal (LIP) region that encode spatial locations. This inhibition parameter is known to be task dependent and can be quantified through fMRI beta regression values (Roggerman et al, 2010). Building on this work, we incorporated this inhibition parameter into a model of neural activation during enumeration which predicted a cost in reaction times when switching from the estimation of large numerosities to the subitization of small numerosities, as well as systematic underestimation when switching from subitizing to estimation. To test our model we designed an enumeration experiment with subitizing and estimation blocks of variable lengths and measured voice-onset reaction times and the enumeration accuracy. Our findings confirmed the predictions suggesting a shared mechanism for subitizing and estimation.The model provides an explanation for the human number sense which cannot be reduced to low-level visual properties like density of texture or spatial frequency (Burr et al, 2008, Current Biology, 18, 857-858) but instead emerges from high-level neural mechanisms which individuate visual-spatial entities.

Meeting abstract presented at VSS 2013


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.