This paper presents a blind source separation process for convolutive mixtures of audio sources. Here undetermined condition that is few microphones than sources has been considered as a mixing model. By an expectation–maximization (EM) algorithm the separation operation is performed in the frequency domain. The T-F masking separation is made use which is a powerful approach for the separation of underdetermined mixtures, especially for the separation of single-channel mixtures. Even under reverberant conditions the process enables to attain a good separation. From the experimental results, separated signals SDR values of speech mixtures is obtained in the range of 7. 5dB while for music mixtures in the range of 2. 9dB. It can be concluded from these values that separation of speech mixtures is better than music mixtures.

1. A. Ozerov and C. Fevotte, "Multichannel nonnegative matrix factorization in convolutive mixtures for audio source separation," IEEE Trans. Audio, Speech, Lang. Process. , vol. 18, no. 3, pp. 550–563, Mar. 2010.
2. R. Olsson and L. Hansen, "Blind separation of more sources than sensors in convolutivemixtures," in Proc. ICAP'06, May 2006, vol. V, pp. 657–660.
3. . O. Yilmaz and S. Rickard, "Blind separation of speech mixtures via time-frequency masking," IEEE Trans. Signal Process. , vol. 52, no. 7, pp. 1830–1847, Jul. 2004
4. E. Vincent, S. Araki, and P. Bofill, "The 2008 signal separation evaluation campaign: A community-based approach to large-scale evaluation," in Proc. ICA'09, 2009 [Online]. Available: http://sisec2008. wiki. irisa. fr/tiki-index. php.
5. H. Sawada, S. Araki, R. Mukai, and S. Makino, "Grouping separated frequency components by estimating propagation model parameters in frequency-domain blind source separation," IEEE Trans. Audio, Speech, Lang. Process. , vol. 15, no. 5, pp. 1592–1604, Jul. 2007.
6. R. Mukai, S. Araki, H. Sawada, and S. Makino, "Evaluation of separation and dereverberation performance in frequency domain blind source separation," Acoust. Sci. Technol. vol. 25, no. 2, pp. 119–126, 2004.
7. Netabayashi, Tomoyuki, "Robustness of the blind source separation with reference against uncertainties of the reference signals".
8. Blind Source Separation Of More Sources Than Mixtures Using Overcomplete Representations, Te-Won Lee, Member, IEEE, Michael S. Lewicki, Mark Girolami, Member, IEEE, and Terrence J. Sejnowski, Senior Member, ieee, IEEE signal processing letters, vol. 6, no. 4, april 1999 87.
9. "Comparison of Blind Source Separation Methods based on Iterative Algorithms J. Rinas, K. D. Kammeyer Department of Communications Engineering, Universitat¨ Bremen, frinas,kammeyerg@ant. uni-bremen. de
10. "Equivalence between Frequency Domain Blind Source Separation and Frequency Domain Adaptive Beamformers", Shoko Araki,Shoji Makino,Ryo Mukai,Hiroshi Saruwatari. NTT Communication Science Laboratories.
11. Equivalence between Frequency Domain Blind Source Separation and Frequency Domain Adaptive Beamformers, Shoko Araki,Shoji Makino,Ryo Mukai,Hiroshi Saruwatari. NTT Communication Science Laboratories.
12. Ngoc Q. K. Duong, Emmanuel Vincent and Remi Gribonval "Under-determined reverberant audio source separation using a full-rank spatial covariance model" IEEE transactions on audio, speech, and language processing.

Blind Source Separation Convolutive Mixtures Undetermined Mixtures Em Algorithm T-f Masking