An Online Blind Source Separation for Convolutive Acoustic Signals in Frequency-Domain
In this paper we propose a scheme for online convolutive blind acoustic source separation in frequency domain. A restriction term of DOA (Direction of Arrival) for sources is added to the cost function of mutual information. Minimizing the modified mutual information in each frequency bin, the source order can be adjusted on line. So the problem of permutation ambiguity can be settled on line, and it does not need the location information of receivers like classical DOA method. Simulation results show that the proposed algorithm has better performance than other existing methods.
KeywordsMutual Information Independent Component Analysis Independent Component Analysis Blind Source Separation Microphone Array
Unable to display preview. Download preview PDF.
- 1.Bell, A., Sejnowski, T.: An information-maximization approach to blind separation and blind deconvolution. Nerualcomputing 7, 1129–1159 (1995)Google Scholar
- 2.Amari, S.: Natural gradient works efficiency in learning. Neuralcomputing 10, 251–276 (1998)Google Scholar
- 4.Jiang, W.D.: Blind source separation of speech signals based on the amplitude correlation of neighbor bins. Journals of Circuits and System, 1–4 (2005)Google Scholar
- 5.Ikram, M.Z., Morgan, D.R.: A beamforming approach to permutation alignment for multichannel frequency-domain blind speech separation. In: Proc. ICASSP, pp. 881–884 (2002)Google Scholar
- 7.Tang, H.W., Qin, X.Z.: Practical Optimization Method, pp. 202–205. Dalian University of Technology Press (2004)Google Scholar
- 9.Convolutive mixturesII (in a virtual room), http://www.ism.ac.jp/~shiro/research/blindsep.html