Display options
Cite
Theiss JD, Bowen JD, Silver MA. Spatial Attention Enhances Crowded Stimulus Encoding Across Modeled Receptive Fields by Increasing Redundancy of Feature Representations. Neural Comput. 2021;34(1):190-218doi: 10.1162/neco_a_01447.
Theiss, J. D., Bowen, J. D., & Silver, M. A. (2021). Spatial Attention Enhances Crowded Stimulus Encoding Across Modeled Receptive Fields by Increasing Redundancy of Feature Representations. Neural computation, 34(1), 190-218. https://doi.org/10.1162/neco_a_01447
Theiss, Justin D, et al. "Spatial Attention Enhances Crowded Stimulus Encoding Across Modeled Receptive Fields by Increasing Redundancy of Feature Representations." Neural computation vol. 34,1 (2021): 190-218. doi: https://doi.org/10.1162/neco_a_01447
Theiss JD, Bowen JD, Silver MA. Spatial Attention Enhances Crowded Stimulus Encoding Across Modeled Receptive Fields by Increasing Redundancy of Feature Representations. Neural Comput. 2021 Dec 15;34(1):190-218. doi: 10.1162/neco_a_01447. PMID: 34710898; PMCID: PMC8693207.
Copy Download .nbib Format: NLM
Share it on

Neural Comput. 2021 Dec 15;34(1):190-218. doi: 10.1162/neco_a_01447.

Spatial Attention Enhances Crowded Stimulus Encoding Across Modeled Receptive Fields by Increasing Redundancy of Feature Representations.

Neural computation

Justin D Theiss, Joel D Bowen, Michael A Silver

Affiliations

  1. University of California, Berkeley, CA 94720, U.S.A. [email protected].
  2. University of California, Berkeley, CA 94720, U.S.A. [email protected].
  3. University of California, Berkeley, CA 94720, U.S.A. [email protected].

PMID: 34710898 PMCID: PMC8693207 DOI: 10.1162/neco_a_01447

Abstract

Any visual system, biological or artificial, must make a trade-off between the number of units used to represent the visual environment and the spatial resolution of the sampling array. Humans and some other animals are able to allocate attention to spatial locations to reconfigure the sampling array of receptive fields (RFs), thereby enhancing the spatial resolution of representations without changing the overall number of sampling units. Here, we examine how representations of visual features in a fully convolutional neural network interact and interfere with each other in an eccentricity-dependent RF pooling array and how these interactions are influenced by dynamic changes in spatial resolution across the array. We study these feature interactions within the framework of visual crowding, a well-characterized perceptual phenomenon in which target objects in the visual periphery that are easily identified in isolation are much more difficult to identify when flanked by similar nearby objects. By separately simulating effects of spatial attention on RF size and on the density of the pooling array, we demonstrate that the increase in RF density due to attention is more beneficial than changes in RF size for enhancing target classification for crowded stimuli. Furthermore, by varying target/flanker spacing, as well as the spatial extent of attention, we find that feature redundancy across RFs has more influence on target classification than the fidelity of the feature representations themselves. Based on these findings, we propose a candidate mechanism by which spatial attention relieves visual crowding through enhanced feature redundancy that is mostly due to increased RF density.

© 2021 Massachusetts Institute of Technology.

References

  1. Nat Rev Neurosci. 2013 Mar;14(3):188-200 - PubMed
  2. Vision Res. 2019 Mar;156:84-95 - PubMed
  3. J Comp Neurol. 1981 Oct 1;201(4):519-39 - PubMed
  4. J Vis. 2014 Oct 16;14(6): - PubMed
  5. Curr Biol. 2019 Jul 8;29(13):2229-2236.e3 - PubMed
  6. J Vis. 2016;16(3):39 - PubMed
  7. Vision Res. 1992 Jul;32(7):1349-57 - PubMed
  8. Annu Rev Neurosci. 2001;24:1193-216 - PubMed
  9. Nature. 1998 Nov 5;396(6706):72-5 - PubMed
  10. Neuron. 2019 Oct 23;104(2):402-411.e4 - PubMed
  11. Proc Natl Acad Sci U S A. 1999 Feb 16;96(4):1663-8 - PubMed
  12. J Vis. 2012 Sep 26;12(10):13 - PubMed
  13. Vision Res. 2011 Jul 1;51(13):1484-525 - PubMed
  14. Trends Cogn Sci. 2011 Apr;15(4):160-8 - PubMed
  15. J Vis. 2013 Mar 22;13(4):17 - PubMed
  16. Nat Neurosci. 2011 Aug 14;14(9):1195-201 - PubMed
  17. J Vis. 2010 May 01;10(5):16 - PubMed
  18. J Exp Psychol Hum Percept Perform. 2014 Jun;40(3):1022-33 - PubMed
  19. Nature. 1970 Apr 11;226(5241):177-8 - PubMed
  20. J Neurosci. 2017 Mar 22;37(12):3386-3401 - PubMed
  21. Vision Res. 2011 May 25;51(10):1117-23 - PubMed
  22. Annu Rev Vis Sci. 2016 Oct 14;2:437-457 - PubMed
  23. J Vis. 2010 Aug 18;10(10):16 - PubMed
  24. Trends Cogn Sci. 2015 Jun;19(6):349-57 - PubMed
  25. Curr Biol. 2018 Feb 5;28(3):R127-R133 - PubMed
  26. J Vis. 2009 Jun 29;9(6):18.1-15 - PubMed
  27. Nat Neurosci. 2006 Nov;9(11):1412-20 - PubMed
  28. J Neurosci. 2009 May 6;29(18):5749-57 - PubMed
  29. Psychol Sci. 2017 Mar;28(3):285-296 - PubMed
  30. Annu Rev Neurosci. 1995;18:193-222 - PubMed
  31. J Neurosci. 1999 Jan 1;19(1):431-41 - PubMed
  32. J Neurosci. 1993 Nov;13(11):4700-19 - PubMed
  33. Trends Cogn Sci. 1999 Apr;3(4):128-135 - PubMed
  34. Neural Netw. 2020 Jun;126:262-274 - PubMed
  35. Nat Neurosci. 2012 Jan 08;15(3):463-9, S1-2 - PubMed
  36. J Neurosci. 1988 Jun;8(6):1831-45 - PubMed
  37. J Neurosci. 1999 Mar 1;19(5):1736-53 - PubMed
  38. Cereb Cortex. 2009 Oct;19(10):2466-78 - PubMed
  39. Percept Psychophys. 2008 Jan;70(1):104-13 - PubMed
  40. Science. 1985 Aug 23;229(4715):782-4 - PubMed
  41. J Vis. 2018 Mar 1;18(3):4 - PubMed
  42. Nat Neurosci. 2006 Sep;9(9):1156-60 - PubMed
  43. Vision Res. 2004 Mar;44(5):501-21 - PubMed
  44. Vision Res. 2008 Feb;48(5):635-54 - PubMed
  45. J Vis. 2015;15(6):5 - PubMed
  46. Neuron. 2009 Jan 29;61(2):168-85 - PubMed
  47. Neuron. 2014 Oct 1;84(1):227-237 - PubMed
  48. J Vis. 2007 Feb 14;7(2):7.1-23 - PubMed
  49. Vision Res. 2008 Jan;48(1):80-95 - PubMed
  50. Front Integr Neurosci. 2014 Sep 25;8:73 - PubMed
  51. J Neurosci. 2014 Aug 6;34(32):10465-74 - PubMed
  52. PLoS Comput Biol. 2010 Jan 22;6(1):e1000646 - PubMed

Publication Types