Gaussian Derivative Model
The Gaussian derivative model of cortical receptive fields was first proposed by Young in 1986
[Young 1986].
- Gaussian derivative functions are real valued functions.
- In 1D they can be characterized as the product of a polynomial (generalized Hermite
polynomial) and a Gaussian.
These are plots of Gaussian derivative functions made up of the second and third derivatives
of a Gaussian (left and right respectively) along the x-axis and a Gaussian along the y-axis.
As was the case with the receptive fields, these functions have some nominal activity (grey)
which inceases (white) and decreases (black) in local, oriented bands.
