Color has become a premier model system for understanding how information is processed by neural circuits, and for investigating the relationships among genes, neural circuits, and perception. Both the physical stimulus for color and the perceptual output experienced as color are quite well characterized, but the neural mechanisms that underlie the transformation from stimulus to perception are incompletely understood. The past several years have seen important scientific and technical advances that are changing our understanding of these mechanisms. Here, and in the accompanying minisymposium, we review the latest findings and hypotheses regarding color computations in the retina, primary visual cortex, and higher-order visual areas, focusing on non-human primates, a model of human color vision.In trichromatic primates, including humans and Old World monkeys, there are three types of cone photoreceptors that are responsible for color vision ( Fig. 1 A, B). The cone classes are called L, M, and S because of their spectral-sensitivity peaks, which lie in the long-, middle-, and short-wavelength regions of the visible spectrum. These labels replace the misleading terms "red," "green," and "blue." Two physically distinct stimuli appear as different colors only if they produce different relative activations in at least two cone types; conversely, any pair of physically distinct stimuli that activate the cone types in the same relative amount appear the same, like the two yellows shown in Figure 1C. While photoreceptor responses are easily computed from the spectral distribution of the stimulus, there is no straightforward relationship between photoreceptor response and color (Hofer et al., 2005a;Shevell and Kingdom, 2008). The multitude of color phenomena, including color afterimages, color assimilation, neon-color spreading, color constancy, and colored shadows, is compelling because in many cases two physically identical stimuli are made to appear different colors, or two physically different stimuli are made to appear the same simply by changing the spatial or temporal context (Fig. 2). A full description of the neural machinery for color should account for these observations, as well as more cognitive phenomena involving the relationship between experience, language, memory, emotion and color. The neural basis of color has been reviewed previously from a range of perspectives (Gegenfurtner, 2003;Gegenfurtner and Kiper, 2003;Lennie and Movshon, 2005;Sincich and Horton, 2005;Solomon and Lennie, 2007;Conway, 2009;Dobkins, 2009;Jacobs and Nathans, 2009;Stockman and Brainard, 2010). Here we focus on advances and pressing questions regarding the mechanisms of color in retina, striate cortex, and extrastriate cortex of non-human primates, although we note that other species are emerging as excellent model systems of color processing (Lotto and Chittka, 2005;Van Hooser and Nelson, 2006;Osorio and Vorobyev, 2008;Borst, 2009;Johnson et al., 2010;Srinivasan, 2010).
Retinal mechanismsA single cone by itself is color blind b...