A neural field model for color perception unifying assimilation and contrast

Abstract : We propose a neural field model of color perception in context, for the visual area V1 in the cortex. This model reconciles into a common framework two opposing perceptual phenomena, simultaneous contrast and chromatic assimilation. Previous works showed that they act simultaneously, and can produce larger shifts in color matching when acting in synergy with a spatial pattern. At some point in an image,the color perceptually seems more similar to that of the adjacent locations, while being more dissimilar from that of remote neighbors. The influence of neighbors hence reverses its nature above some characteristic scale. Our model fully exploits the balance between attraction and repulsion in color space, combined at small or large scales in physical space. For that purpose we rely on the opponent color theory introduced by Hering, and suppose a hypercolumnar structure coding for colors. At some neural mass, the pointwise influence of neighbors is spatially integrated to obtain the final effect that we call a color sensation. Alongside this neural field model, we describe the search for a color match in asymmetric matching experiments as a mathematical projector. We validate it by fitting the parameters of the model to data from (Monnier and Shevell, 2004) and (Monnier, 2008) and our own data. All the results show that we are able to explain the nonlinear behavior of the observed shifts along one or two dimensions in color space, which cannot be done using a simple linear model.
Document type :
Preprints, Working Papers, ...
Liste complète des métadonnées

https://hal.inria.fr/hal-01909354
Contributor : Romain Veltz <>
Submitted on : Wednesday, October 31, 2018 - 9:43:15 AM
Last modification on : Friday, April 19, 2019 - 4:54:57 PM

Links full text

Identifiers

Citation

Anna Song, Olivier Faugeras, Romain Veltz. A neural field model for color perception unifying assimilation and contrast. 2018. ⟨hal-01909354⟩

Share

Metrics

Record views

55