CaltechAUTHORS
  A Caltech Library Service

Temporal dynamics of the neural representation of hue and luminance contrast

Hermann, Katherine L. and Singh, Shridhar R. and Rosenthal, Isabelle A. and Pantazis, Dimitrios and Conway, Bevil R. (2020) Temporal dynamics of the neural representation of hue and luminance contrast. . (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20200626-103613101

[img] PDF (November 14, 2020) - Submitted Version
Creative Commons Public Domain Dedication.

4Mb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20200626-103613101

Abstract

Hue and luminance contrast are the most basic visual features, emerging in early layers of convolutional neural networks trained to perform object categorization. In human vision, the timing of the neural computations that extract these features, and the extent to which they are determined by the same or separate neural circuits, is unknown. We addressed these questions using multivariate analyses of human brain responses measured with magnetoencephalography. We report four discoveries. First, it was possible to decode hue tolerant to changes in luminance contrast, and luminance contrast tolerant to changes in hue, consistent with the existence of separable neural mechanisms for these features. Second, the decoding time course for luminance contrast peaked 16-24 ms before hue and showed a more prominent secondary peak corresponding to decoding of stimulus cessation. These results support the idea that the brain uses luminance contrast as an updating signal to separate events within the constant stream of visual information. Third, neural representations of hue generalized to a greater extent across time, providing a neural correlate of the preeminence of hue over luminance contrast in perceptual grouping and memory. Finally, decoding of luminance contrast was more variable across participants for hues associated with daylight (orange and blue) than for anti-daylight (green and pink), suggesting that color-constancy mechanisms reflect individual differences in assumptions about natural lighting.


Item Type:Report or Paper (Discussion Paper)
Related URLs:
URLURL TypeDescription
https://doi.org/10.1101/2020.06.17.155713DOIDiscussion Paper
ORCID:
AuthorORCID
Pantazis, Dimitrios0000-0002-2146-9065
Conway, Bevil R.0000-0001-7715-9253
Alternate Title:Temporal dynamics of the neural mechanisms for encoding hue and luminance contrast uncovered by magnetoencephalography
Additional Information:The copyright holder for this preprint is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license. We thank Chris Baker, Susan Wardle, Santani Teng, and Qasim Zaidi for helpful discussions. This research was supported in part by the Intramural Research Program of the National Institutes of Health, National Eye Institute, and grant R01 EY-0233223 to BRC. Author contributions: KH and BRC conceived and designed the experiments; KH and IR collected the data; KH, IR, and SS analyzed the data; IR, SS and BRC made the figures; DP provided expertise and resources for MEG; BRC supervised the work and wrote the paper. Data Availability: All raw datasets analyzed in the current study will be available in the NEICOMMONS repository. Code Availability: The procedures and code used to produce all the figures, and all statistical analyses, in the current study will be available in the NEICOMMONS repository.
Funders:
Funding AgencyGrant Number
NIHR01 EY-0233223
Record Number:CaltechAUTHORS:20200626-103613101
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20200626-103613101
Official Citation:Temporal dynamics of the neural representation of hue and luminance contrast. Katherine Hermann, Shridhar Singh, Isabelle Rosenthal, Dimitrios Pantazis, Bevil Conway. bioRxiv 2020.06.17.155713; doi: https://doi.org/10.1101/2020.06.17.155713
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:104078
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:26 Jun 2020 19:56
Last Modified:16 Nov 2020 22:07

Repository Staff Only: item control page