September 2015
Volume 15, Issue 12
Vision Sciences Society Annual Meeting Abstract  |   September 2015
Shared noise variability facilitates discrimination of natural images in V4 population
Author Affiliations
  • Shaobo Guan
    Department of Neuroscience, Brown University
  • Ruobing Xia
    Department of Neuroscience, Brown University
  • David Sheinberg
    Department of Neuroscience, Brown University
Journal of Vision September 2015, Vol.15, 1097. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Shaobo Guan, Ruobing Xia, David Sheinberg; Shared noise variability facilitates discrimination of natural images in V4 population. Journal of Vision 2015;15(12):1097.

      Download citation file:

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

  • Supplements

The visual system is usually simplied as a hierarchical feed-forward filtering system containing independently functioning neurons. However, interactions among visual neurons exist in neural population. The pairwise shared noise variability of spike counts, or noise correlation, is an easily accessible and potentially powerful measure of such interactions. About the role of noise correlation in a population code, the literature includes multiple, often conflicting, perspectives. This is partly due to how noise correlation is experimentally quantified (Cohen & Kohn, 2011), and partly to how its impact is evaluated. In the present study, we attempted to address these discrepancies and to more objectively assess noise correlation and its role in visual coding. We used a 32-channel semi-chronic microdrive array to record spiking activity from the foveal region of macaque area V4 in a passive viewing task. Our study differed from previous investigations in three primary respects: 1) We used complex natural images as visual stimuli, instead of artificial patterns with limited variability, to better approximate naturalistic visual experience; 2) we quantified noise correlation independently for each stimulus rather than using the mean correlation, so as to preserve the stimulus modulation of correlation that is informative for coding (Ponce-Alvarez, Thiele, Albright, Stoner, & Deco, 2013); 3) we evaluated the coding performance using the likelihood ratio of a multi-variant gaussian model, a decoder approach that provides advantages over direct estimation of information given limited data (Moreno-Bote et al., 2014). Our preliminary results suggest that the pattern of noise correlation improves discrimination of natural stimuli, providing the first evidence of synergistic processing with neurons exhibiting shared noise variability.

Meeting abstract presented at VSS 2015


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.