Two Pairwise Iterative Schemes For High Dimensional Blind Source Separation

TL;DR

This paper introduces two pairwise iterative schemes utilizing a novel divergence measure to efficiently perform blind source separation in high-dimensional settings, outperforming existing algorithms in speed and accuracy.

Contribution

The paper proposes two innovative pairwise iterative schemes based on Convex Cauchy-Schwarz Divergence for high-dimensional blind source separation, improving efficiency and performance.

Findings

Schemes enable fast, efficient source demixing in high dimensions

Performance surpasses FastICA, RobustICA, and CICA in metrics

Demonstrated effectiveness on real-world high-dimensional data

Abstract

This paper addresses the high dimensionality problem in blind source separation (BSS), where the number of sources is greater than two. Two pairwise iterative schemes are proposed to tackle this high dimensionality problem. The two pairwise schemes realize nonparametric independent component analysis (ICA) algorithms based on a new high-performance Convex CauchySchwarz Divergence (CCSDIV). These two schemes enable fast and efficient demixing of sources in real-world high dimensional source applications. Finally, the performance superiority of the proposed schemes is demonstrated in metric-comparison with FastICA, RobustICA, convex ICA (CICA), and other leading existing algorithms.