Abstract
Binding theory is used to understand how the activity of distributed cortical cells, driven by different parts or aspects of the same image, may be integrated into a coherent percept. It has found widespread application to feature integration and figure/ground segmentation problems. Much evidence points to a type of oscillatory synchronization as a binding mechanism. Recently we showed how ideas from binding theory could be used to understand the fading and fragmentation of stabilized images and the melting of equiluminant images (Billock & Tsou, TINS, 2004; JCNS, 2004). Here, I consider some additional applications of binding theory to color psychophysics: (1) It is possible to use oscillatory synchronization to recode sensory information from two or more channels; for example, it is relatively straightforward to use oscillatory synchronization to create a chromatic brightness channel using the compromise frequency of nonlinear coupled hue and luminance channels. (2) The same principle can be used to recode brightness and hue mechanisms, to compute chromatic saturation. Both this saturation signal and the brightness signal obey a power law scaling that is an emergent bonus of the nonlinear dynamic interactions between the channels. (3) It has long been a mystery why both parvo- and magno-driven mechanisms converge on nearly identical spectral sensitivities. Synchronization of two neural mechanisms is better behaved when the correlation between the two channels is high; it may be easier to synchronize luminance inputs from different neural mechanisms if this sensitivity-induced-correlation is present. Even if direct addition of luminance mechanism is employed, a high correlation leading to synchronization eliminates two sources of contrast demodulation. Treating oscillatory binding as a generic neural mechanism rather than a special case may be quite useful in understanding psychophysical phenomena and neural processing.