Blind Source Separation of Many Signals in the Frequency Domain

This paper describes the frequency-domain blind source separation (BSS) of convolutively mixed acoustic signals using independent component analysis (ICA). The most critical issue related to frequency domain BSS is the permutation problem. This paper presents two methods for solving this problem. Both methods are based on the clustering of information derived from a separation matrix obtained by ICA. The first method is based on direction of arrival (DOA) clustering. This approach is intuitive and easy to understand. The second method is based on normalized basis vector clustering. This method is less intuitive than the DOA based method, but it has several advantages. First, it does not need sensor array geometry information. Secondly, it can fully utilize the information contained in the separation matrix, since the clustering is performed in high-dimensional space. Experimental results show that our methods realize BSS in various situations such as the separation of many speech signals located in a 3-dimensional space, and the extraction of primary sound sources surrounded by many background interferences

[1]  Andreas Ziehe,et al.  An approach to blind source separation based on temporal structure of speech signals , 2001, Neurocomputing.

[2]  Scott C. Douglas,et al.  Convolutive blind separation of speech mixtures using the natural gradient , 2003, Speech Commun..

[3]  Terrence J. Sejnowski,et al.  An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[4]  Nobuhiko Kitawaki,et al.  A combined approach of array processing and independent component analysis for blind separation of acoustic signals , 2001, 2001 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings (Cat. No.01CH37221).

[5]  Hiroshi Sawada,et al.  A NOVEL BLIND SOURCE SEPARATION METHOD WITH OBSERVATION VECTOR CLUSTERING , 2005 .

[6]  Aapo Hyvärinen,et al.  A Fast Fixed-Point Algorithm for Independent Component Analysis of Complex Valued Signals , 2000, Int. J. Neural Syst..

[7]  David G. Stork,et al.  Pattern Classification , 1973 .

[8]  Hiroshi Sawada,et al.  Frequency-Domain Blind Source Separation , 2007, Blind Speech Separation.

[9]  Hiroshi Sawada,et al.  Blind extraction of a dominant source from mixtures of many sources using ICA and time-frequency masking , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[10]  Hiroshi Sawada,et al.  Frequency Domain Blind Source Separation for Many Speech Signals , 2004, ICA.

[11]  Nobuhiko Kitawaki,et al.  Combined approach of array processing and independent component analysis for blind separation of acoustic signals , 2003, IEEE Trans. Speech Audio Process..

[12]  Kiyohiro Shikano,et al.  Blind Source Separation Combining Independent Component Analysis and Beamforming , 2003, EURASIP J. Adv. Signal Process..

[13]  Hiroshi Sawada,et al.  Polar coordinate based nonlinear function for frequency-domain blind source separation , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[14]  Paris Smaragdis,et al.  Blind separation of convolved mixtures in the frequency domain , 1998, Neurocomputing.

[15]  Hiroshi Sawada,et al.  A spatio-temporal fastICA algorithm for separating convolutive mixtures , 2005, Proceedings. (ICASSP '05). IEEE International Conference on Acoustics, Speech, and Signal Processing, 2005..

[16]  David G. Stork,et al.  Pattern classification, 2nd Edition , 2000 .

[17]  Hiroshi Sawada,et al.  A robust and precise method for solving the permutation problem of frequency-domain blind source separation , 2004, IEEE Transactions on Speech and Audio Processing.

[18]  K. Matsuoka,et al.  Minimal distortion principle for blind source separation , 2002, Proceedings of the 41st SICE Annual Conference. SICE 2002..