October 2020
Volume 20, Issue 11
Open Access
Vision Sciences Society Annual Meeting Abstract  |   October 2020
Anisotropic representation of orientation by convolutional neural networks
Author Affiliations & Notes
  • Margaret Henderson
    Neurosciences Graduate Program, UC San Diego
  • John Serences
    Neurosciences Graduate Program, UC San Diego
    Department of Psychology, UC San Diego
    Kavli Institute for Brain and Mind, UC San Diego
  • Footnotes
    Acknowledgements  Funding provided by NEI R01-EY025872 to J.T.S. and a UCSD Institute for Neural Computation predoctoral fellowship awarded to M.M.H.
Journal of Vision October 2020, Vol.20, 224. doi:https://doi.org/10.1167/jov.20.11.224
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Margaret Henderson, John Serences; Anisotropic representation of orientation by convolutional neural networks. Journal of Vision 2020;20(11):224. https://doi.org/10.1167/jov.20.11.224.

      Download citation file:


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

      ×
  • Supplements
Abstract

Visual representations learned by convolutional neural networks (CNNs) share some similarity in representational structure to neural representations in the primate ventral visual stream (e.g. Yamins et al., 2014). However, the organization of low-level feature representations by CNNs has not been extensively characterized. Understanding whether CNNs develop idiosyncrasies that mimic the properties of the primate visual system is important for developing models that can inform our understanding of the brain. Additionally, because many aspects of CNN representations are acquired through training, examining feature representations of CNNs is a useful tool for determining which properties of the primate brain might be innate and which are likely to be acquired through experience. Here, we focus on orientation perception, a well-understood aspect of the primate visual system. We asked whether convolutional neural networks trained to perform object recognition on a natural image database would exhibit an “oblique effect” such that cardinal (vertical and horizontal) orientations are represented with higher precision than oblique (diagonal) orientations, as has been measured in the brain and behavior of primates. We obtained activation patterns from a pre-trained VGG-16 network (Simonyan & Zisserman, 2014) presented with oriented grating stimuli, and used a Euclidean distance metric to measure the discriminability between patterns corresponding to different pairs of orientations. In agreement with human perception, we find that orientation discriminability generally peaked around the cardinal orientations. This effect emerged at middle layers of the VGG-16 network. Its magnitude increased with stimulus spatial frequency, but decreased with stimulus uncertainty. We also trained networks from scratch using images from the ImageNet database (Deng et al., 2009) that had been rotated by varying increments. Overall, our findings suggest that cardinality effects in human visual perception are not dependent on a hard-wired anatomical bias, but can instead emerge through experience with the statistics of natural images.

×
×

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.

×