DEEP Neural Networks
show how the brain is wired
FFDs in the Brain
This is a summary of the significance of FFDs in the brain. Details can be found in [1]-[3] in the Related Articles section.
Evidence suggests FFDs carry out much of the information processing in the brain. This family of neural logic circuits represents the first neural model that is explicit and predictive of known brain phenomena. FFDs show for the first time that a neuron can function as a logic gate, that it is functionally complete, and that inhibition can have a significant role in processing information. The logic circuits provide the foundation for all of the brain’s combinational processing of information.
FFDs generate correlates of known psychophysical phenomena. These results follow directly from the FFD architecture and the minimal cellular capabilities of excitation and inhibition. The networks function dynamically, making their operation consistent with the speed of most brain functions. The networks show that well-known psychophysical phenomena do not require extraordinarily complex brain structures, and that a single network architecture can produce apparently disparate phenomena in different sensory systems.
FFDs also produce correlates of electrophysiological brain phenomena and predict major aspects of the anatomical structure and physiological organization of the neocortex. At the local level, FFDs account for much of the physical structure of the neocortex as well its organization of synaptic connections.
The six outputs of the FFD model of color vision model transform sensory data into neural correlates of color and black and white, and they provide correlates of hue, saturation, and brightness. The color cells make chromatic color distinctions, and the model explains the achromatic characteristics of black and white.
The model’s outputs define a three-dimensional Color Space that provides detailed, quantitative neural explanations of the shapes and color relationshiops of standard color figures such as Newton’s color wheel and Munsell’s color system. The output values are rectangular coordinates of each color point in the figure, and the model’s hue, saturation, and brightness are cylindrical coordinates.
The model also explains the continuous yet discrete nature of color; mutually exclusive colors and colors that can be perceived together; color mixing; the perceptions of unique blue, green, and yellow at specific wavelengths; the Bezold-Brücke hue shift; the additivity “failure” of brightness; and opponent-color cells.
Generating Apparently Disparate Phenomena in Different Sensory Systems in the Same Way [2]
FFDs generate the main known phenomena of olfaction and produce identical color phenomena. Because the phenomena are experienced in different sensory systems, they may not have been previously recognized as the same. The FFD explanation for both olfaction and color is the same. It is significant that in the cases where the brain “fails,” FFDs fail in exactly the same way.
Perceptual Independence of Stimulus Intensity. Under most ordinary circumstances, the brain can identify an odor or color independently of its strength. (A skunk’s odor smells like a skunk regardless of its concentration, and a yellow banana appears to be yellow independently of the illuminant’s intensity.)
Perceptual Dependence on Stimulus Intensity. On the other hand, the perceptions of some odors and colors vary with the odorant concentration or photostimulus intensity. For colors this is known as the Bezold-Brücke hue shift. In this case, the FFD model not only predicts this can happen, but it also predicts which colors will change and in what ways. Specifically, orange and greenish yellow both appear yellower at higher photostimulus intensities; violet and greenish blue appear bluer.
Perceptual Complexity. The fuzzy logic of FFDs discriminates individual odors in a mixture of odorants, even odorants that activate nearly identical sets of receptor types. Perceptions of mixtures provide especially good tests of any model because mixtures can elicit complex patterns of sensory receptor responses that result in numerous varied and complex perceptions that can differ markedly from the perceptions of the mixture’s components alone. Odorant mixtures have been found to have four general properties. The FFD models produce them and correctly predict identical phenomena for color perception.
1. Perceived complexity in a single odorant or photostimulus can be greater than in a mixture of the odorant or photostimulus with several others.
2. No more than three or four odors or colors (including black and white) can be discriminated in mixtures.
3. Neither the chemical complexity of an odorant nor the spectral complexity of a photostimulus is correlated with the perceived complexity.
4. Perceived complexity is not additive when odorants or photostimuli are mixed.
Local Properties of the Neocortex’s Physiology and Anatomy [3]
FFDs account for significant local properties of the neocortex’s anatomical structure and physiological organization, including its laminar configuration, the organization in distinct layers of cells, five cellular layers, axons extending mainly parallel or perpendicular to the layers, axons in the outer layers mainly parallel to the layers, projections from one part of the neocortex to another coming mainly from neurons in layers II and III, projections to subcortical regions mainly in the lower layers, and inhibitory cells located in all cellular layers and constituting 20 to 25 percent of the neurons.
The architecture also predicts local electrophysiological phenomena of the neocortex, including information processed in small columns of neurons connected across the layers, information transformed in each layer while passing through the columns in series and in parallel, and active regions surrounded by neurons that are suppressed below the resting level.