Self-stabilized gradient algorithms for blind source separation with orthogonality constraints

Authors:
S. C. Douglas
Affiliations:
Dept. of Electr. Eng., Southern Methodist Univ., Dallas, TX, USA
Venue:
IEEE Transactions on Neural Networks
Year:
2000

Citing 0
Cited 7

Gradient Adaptive Algorithms for Contrast-Based Blind Deconvolution

Journal of VLSI Signal Processing Systems
On a Class of Orthonormal Algorithms for Principal and Minor Subspace Tracking

Journal of VLSI Signal Processing Systems
Gradient Adaptive Paraunitary Filter Banks for Spatio-Temporal Subspace Analysis and Multichannel Blind Deconvolution

Journal of VLSI Signal Processing Systems
Accurate estimation of ICA weight matrix by implicit constraint imposition using lie group

IEEE Transactions on Neural Networks
Geometrical methods for non-negative ICA: Manifolds, Lie groups and toral subalgebras

Neurocomputing
Letters: A fast NPCA algorithm for online blind source separation

Neurocomputing
An ICA learning algorithm utilizing geodesic approach

ISNN'06 Proceedings of the Third international conference on Advances in Neural Networks - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

Developments in self-stabilized algorithms for gradient adaptation of orthonormal matrices have resulted in simple but powerful principal and minor subspace analysis methods. We extend these ideas to develop algorithms for instantaneous prewhitened blind separation of homogeneous signal mixtures. Our algorithms are proven to be self-stabilizing to the Stiefel manifold of orthonormal matrices, such that the rows of the adaptive demixing matrix do not need to be periodically reorthonormalized. Several algorithm forms are developed, including those that are equivariant with respect to the prewhitened mixing matrix. Simulations verify the excellent numerical properties of the proposed methods for the blind source separation task.