Complex-valued Independent Component Analysis of Natural Images |
|||
Valero Laparra, Michael Gutmann,
Jesús Malo and Aapo
Hyvarinen Lect. Not. Comp. Sci. (Proc. ICANN 2011), Vol. 6792, pp. 213-220, 2011 |
|||
|
|||
Abstract
Linear
independent component analysis (ICA) learns simple cell receptive
fields from natural images. Here, we show that linear complex-valued
ICA learns complex cell properties from Fourier transformed natural
images, i.e. two Gabor-like filters with quadrature phase relationship.
Conventional methods for complex-valued ICA assume that the phases of
the output signals have uniform distribution. We relax this assumption
by modeling of the phase information of the output sources in the
complex-valued ICA estimation. The resulting model of phases shows that
the distributions are often far from uniform, and the shapes of the
Gabor filters are also changed. |
|||
|
|||
Key Words: Complex
Independent Components Analysis, Natural Image Statistics, Modeling
Fourier phase distribution, Quadrature Phase Receptive Fields
References: 18 |
|||
|
|||
|
|