Abstract
Brightness depends on the spatial scale of the inducers. For example using a disk-and-ring stimulus, Ried & Shapley (1988) showed that a larger width of ring (the inducer) induced strong brightness contrast to the disk (the inducee) compared to a narrower ring, and brightness of the disk in a narrower ring was instead affected much by the luminance of the background field (assimilation). From these results, the authors argued that brightness depends not only on the local contrast between the inducer (ring) and the inducee (disk) but also on long-range interaction among surfaces (disk and background) that produces assimilation, counteracting contrast effect. The spatial scale of the inducer affects relative contributions of contrast and assimilation on brightness computation. It was further argued that the local contrast is readily attributed to the retinal lateral inhibition (antagonistic center-surround processing), but the origin of long-range interaction was deemed uncertain and perhaps post-retinal. However, recent neurophysiological studies discovered that the retinal inhibitory feedback interneurons, horizontal cells (Packer & Dacey, 2005) and some amacrine cell types (Kolb, 1997), manifest spatially extended receptive-fields (wide RFs). We simulated Ried & Shapley's experiment in two different biophysical retinal model platforms (van Hateren, 2007; Wilson, 1997) in consideration of wide RF of interneurons and obtained results that qualitatively replicate their behavioral data. To the best of our knowledge this would be the first evidence that the local contrast and long-range surface interaction share the same neural locus and that brightness assimilation may already be taking place at retinal level.
This work was supported by the European Research Council, Starting Grant ref. 306337, by the Spanish government, grant ref. TIN2012-38112, and by the Icrea Academia Award.