Abstract

Separating one source from a mixture of sources is a problem, normally observed with parties. Here the sources may be all speech signals or one is speech and the other is music. To have a better understanding of speech, needy to separate actual signal. This can be done by using blind source separation technique. It is hard to extract an interesting conversation from the background noisy crowd. Speech mixture is despoiled by the surrounding noise, interferences and additional speakers. Here an attempt for solving this separation problem, i. e. extracting one or more speech signals from a speech mixture. To eliminate or reduce the noise in speech signal in speech mixture is done by using wavelets. The wavelet output speech mixture processes for source separation by using, two techniques ICA and binary T-F masking. This separation technique is likewise applicable to segregate speech signal under reverberant conditions.

References

• M. Akay ,time frequency and wavelets in biomedical signal processing . Piiscataway, NJ: IEEE Press, 1998 ,pp. 1. Alexis favot and Markus Erne," improved cocktail-party processing", proc. Of the 9th Int conference on digital audio effects(DAFx-06),Montreal, Canada September 18-20,2006
• Kenneth E . Hild and David Pinto,"Convolutive blind source separation by minimizing matual information between segments of signals", IEEE transactions on circuits and systems ,regular papers , vol. 52, No 10, October 2005
• Robi polikar, ''the engineer's ultimate guide to wavelet analysis,'' hosted by Rowan university , college of engineering web servers , last major updates January 2001.
• Dr. Michael lewicki ,"Michael lewicki computational perception and scene analysis cource," webserver.
• David L. donoho ,"Denoising via soft thresholding . IEEE Transactions on information theory , 41:613-627, may 1995.
• Subband representation ", IEEE Transactions on speech and audio processing, Vol. 9. no. 5 123-135.
• David L. Donoho and lain M. Johnstone,"Idel spatial adaption via wavelet shrinkage. Biometrika'', 81:425-455, September 1994.
• E. Visser and T. W. Lee, "Speech enhancement using blind source separation and two channel energy based speaker detection,"IEEE Int. Conf. On Acoust. Speech and Signal Process. vol. 1, pp. 884–887, April 2003.
• Tomasz Rutkowski and Andrzej Cichocki,"Speech extraction from interferences in real environment using bank of filters and blind source separation", IEEE, Neural networks, 2002.
• William Addison and Stephen Roberts, "Blind Source Separation with Non Stationary Mixing Using Wavelets, "Pattern Analysis Research Group, the University of Liverpool, 2006.