Blind Source Separation

Blind Source Separation


Blind Source Separation Using Temporal Predictability

The method described in the paper, "Blind Source Separation Using Temporal Predictability" (Neural Computation, 13(7), July, 2001, View Abstract).

The method is summarised here, and here is the MatLab code. A key feature of this code is that it executes source separation of 1D signals using a fast eigenvalue routine.


Independent Component Analysis Matlab Code

MatLab code for using independent component analysis (ICA) can be downloaded from here. Summary information about this code can be viewed from the README file. This code is based on the method described in Bell and Sejnowski's paper "An Information-Maximization Approach to Blind Separation and Blind Deconvolution" (Neural Computation, 7, 1129-1159, 1995). Key features of this code are that 1) it uses a conjugate gradient method, and does not therefore require a learning rate parameter, 2) it executes spatial or temporal or spatiotemporal ICA of 2D images, 3) it has the option to specify a skewed pdf. The above MatLab code can be downloaded from here.

Code for non-square unmixing matrices (with more mixtures than sources) is also avilable from here. The paper that describes the method can be found here ("Undercomplete Independent Component Analysis for Signal Separation and Dimension Reduction").