Purchase this article with an account.
Christopher Shooner, Kathy T. Mullen; Enhancement of luminance sensitivity by chromatic pedestals - a model of nonlinear parvocellular luminance processing. Journal of Vision 2020;20(11):856. doi: https://doi.org/10.1167/jov.20.11.856.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual interactions between red-green (RG) color and luminance contrast offer insight into how these attributes are jointly encoded by parvocellular neurons of the LGN and their cortical targets. Here we show that the enhancement of luminance sensitivity by RG chromatic pedestals can be accounted for by a model in which luminance is computed cortically from the rectified outputs of parvocellular-like units. We used uniform luminance patch stimuli (4 deg diameter, 300 ms duration) presented foveally, superimposed on RG, luminance and a range of combined color-luminance pedestals of varying intensities. In a 2-interval task, subjects (n=3) identified which stimulus had the luminance increment. We also tested RG pedestals peripherally (8 deg horizontally) using an equivalent spatial 2AFC task. Every pedestal tested reduced luminance thresholds at some intensity. The form of facilitation varied with pedestal chromaticity, with chromatic pedestals having an effect at lower pedestal contrasts. Our model predicts this chromatic tuning from the cone-opponency of parvocellular-like units. The outputs of L- and M-cone center units are rectified then summed in a luminance mechanism with an expansive output nonlinearity. Rectification prevents cancellation of L-M opponent signals, allowing RG pedestals to interact with luminance signals at the later nonlinear stage. Notably, the effects pedestals had on luminance sensitivity were not predictable from their visibility. Foveally, chromatic pedestals were visible at contrasts much lower than those that influenced luminance, unlike luminance pedestals, which had an effect near their detection threshold. In the periphery, RG pedestal detection thresholds increased by a factor of 4 or more, but the site of facilitation changed less, such that RG pedestals had an effect near their detection threshold. This dissociation suggests that the mechanisms mediating pedestal facilitation, though chromatically tuned, are distinct from those limiting chromatic detection.
This PDF is available to Subscribers Only