A fast blind source separation algorithm for binaural hearing aids based on frequency bin selection

In binaural signal processing, due to the large distance between two ears, spatial aliasing occurs at the high frequencies. Therefore, the fast algorithms based on frequency bin selection usually limit the frequency range to below 1kHz to avoid dealing with the aliasing problem. In this paper, a fast Blind Source Separation (BSS) algorithm based on frequency bin selection without limiting the range of frequency bin selection was proposed. Efficient method was used to estimate the propagation model parameters to solve the inaccurate delay problem and permutation ambiguity. Besides, we used outlier detection method to further remove frequency bins with poor separation performance after the first selection, which ensures the accuracy of the normalized attenuation and delay matrices. Simulation results show that the proposed algorithm reduces computational complexity and improves separation performance compared with limited range frequency bin selection BSS.

[1]  Rémi Gribonval,et al.  Performance measurement in blind audio source separation , 2006, IEEE Transactions on Audio, Speech, and Language Processing.

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

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

[4]  Christian Jutten,et al.  Blind separation of sources, part I: An adaptive algorithm based on neuromimetic architecture , 1991, Signal Process..

[5]  Makoto Yamada,et al.  A semi-blind source separation method with a less amount of computation suitable for tiny DSP modules , 2009, INTERSPEECH.

[6]  Kazuya Takeda,et al.  Improved Method of Blind Speech Separation with Low Computational Complexity , 2011 .

[7]  Kiyohiro Shikano,et al.  Blind source separation based on a fast-convergence algorithm combining ICA and beamforming , 2006, IEEE Transactions on Audio, Speech, and Language Processing.

[8]  Hiroshi Sawada,et al.  Grouping Separated Frequency Components by Estimating Propagation Model Parameters in Frequency-Domain Blind Source Separation , 2007, IEEE Transactions on Audio, Speech, and Language Processing.

[9]  Scott C. Douglas,et al.  Scaled Natural Gradient Algorithms for Instantaneous and Convolutive Blind Source Separation , 2007, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.

[10]  Erkki Oja,et al.  Independent Component Analysis , 2001 .

[11]  S.C. Douglas,et al.  Multichannel blind deconvolution and equalization using the natural gradient , 1997, First IEEE Signal Processing Workshop on Signal Processing Advances in Wireless Communications.

[12]  Yi Hu,et al.  Evaluation of Objective Quality Measures for Speech Enhancement , 2008, IEEE Transactions on Audio, Speech, and Language Processing.

[13]  Diego H. Milone,et al.  Correlated Postfiltering and Mutual Information in Pseudoanechoic Model Based Blind Source Separation , 2011, J. Signal Process. Syst..

[14]  Mohammad Hossein Kahaei,et al.  High-speed separation of anechoic mixtures of speech signals using fusion of iterative and closed-forms separation method , 2015 .