A simple closed-form solution for overdetermined blind separation of locally sparse quasi-stationary sources
Authors:
- 1 March 2012
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 2409-2412
Abstract
We consider the scenario of an unknown overdetermined instantaneous mixture of quasi-stationary sources. Blind source separation (BSS) under this scenario has drawn much attention, motivated by applications such as speech and audio separation. The ideas in the existing BSS works often focus on exploiting the time-varying statistics characteristics of quasi-stationary sources, through various kinds of formulations and optimization methods. In this paper, we are interested in further assuming that the sources exhibit some form of local sparsity, which is generally satisfied in speech. By exploiting this additional assumption, we show that there is a simple closed-form solution for the BSS problem. Simulation results based on real speech show that the proposed closed-form algorithm is computationally much lower than some existing BSS algorithms, while delivering a promising mean-square-error performance.Keywords
This publication has 10 references indexed in Scilit:
- A Simplex Volume Maximization Framework for Hyperspectral Endmember ExtractionIEEE Transactions on Geoscience and Remote Sensing, 2011
- Convex analysis for non-negative blind source separation with application in imaging2009
- Batch and Adaptive PARAFAC-Based Blind Separation of Convolutive Speech MixturesIEEE Transactions on Audio, Speech, and Language Processing, 2009
- Underdetermined Blind Separation of Nondisjoint Sources in the Time-Frequency DomainIEEE Transactions on Signal Processing, 2007
- Separating More Sources Than Sensors Using Time-Frequency DistributionsEURASIP Journal on Advances in Signal Processing, 2005
- A frequency domain method for blind source separation of convolutive audio mixturesIEEE Transactions on Speech and Audio Processing, 2005
- Blind Separation of Speech Mixtures via Time-Frequency MaskingIEEE Transactions on Signal Processing, 2004
- Blind separation of instantaneous mixtures of nonstationary sourcesIEEE Transactions on Signal Processing, 2001
- Blind Source Separation by Sparse Decomposition in a Signal DictionaryNeural Computation, 2001
- Parallel factor analysis in sensor array processingIEEE Transactions on Signal Processing, 2000