Webster and Mollon  did some experiments recently on changes in colour appearance following post-receptoral adaptation. The color in the cortex is represented by two statistically independent channels which in our case corresponding to the two neurons. After adaptation to a color hue of angle , the appearance of the color hue at is changed. They measured explicitly the angle of the appearing color hue rotated rotated from the real one (). In Figure 7, the theory (Equation 20) and the experiment data are compared. It is amazing to see how well the theory matched the experiments given that the theoretical curve has only one parameter.
Figure 7: Comparison with experiment for color adaptation of human vision. The diamond symbols are the experimental data. The theoretical curve (solid line) has only one parameter which is the strength of decorrelation feedback.
The same equation is also derived by Atick et al. , although they also used decorrelation, but like Linsker [16,17], their approach emphaseses feedforward path rather than feedback path. Our theory of visual cortex emphaseses the function and adaptation of the connections within cortex. As is shown in the paper that our theory not only predict the simple color adaptation result but also predict the effects of orientation illusions.