The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지
DOI QR코드

DOI QR Code

Blind Signal Separation Using Eigenvectors as Initial Weights in Delayed Mixtures

지연혼합에서의 초기 값으로 고유벡터를 이용하는 암묵신호분리

  • Published : 2006.01.01
Download PDF
⟨ Previous Next ⟩

Abstract

In this paper. a novel technique to set up the initial weights in BSS of delayed mixtures is proposed. After analyzing Eigendecomposition for the correlation matrix of mixing data. the initial weights are set from the Eigenvectors ith delay information. The Proposed setting of initial weighting method for conventional FDICA technique improved the separation Performance. The computer simulation shows that the Proposed method achieves the improved SIR and faster convergence speed of learning curve.

본 논문에서는 지연혼합에서의 암묵신호분리를 위해 분리행렬의 초기 값을 설정하는 방법을 제안한다. 혼합신호의 상호상관행렬에 대한 고유분리를 분석한 후, 고유벡터의 지연정보를 이용하여 초기 값으로 설정한다. 제안하는 방법을 기존의 주파수영역 독립성분분석 (FDICA: Frequency domain independent component analysis)에 초기 값으로 설정하여 분리 성능을 향상시킨다. 컴퓨터 시뮬레이션을 통해 제안하는 방법이 신호대간섭비 (SIR: Signal to Interference Ratio)가 우수하고 학습곡선의 수렴속도가 개선됨을 보인다.

Keywords

References

  1. T. Ristaniemi and, J. Joutsensalo, 'On the performance of blind source separation in CDMA downlink, Proc. Int. Workshop on Independent Analysis and Signal Separation (ICA '99), 437-441, Aussois, France, 1999
  2. S. Makeig, T. Jung, A.J. Bell, D. Ggahremani and, T.J. Seijnowski, 'Blind separation of auditory eventrelated brain response into independent components,' Proceedings on National Academic Science. USA, 10979-10984, 1997
  3. T.W.Lee, A.J.Bell and, R.Orglmeister, 'Blind source separation of real world signals,' Neural Networks, International Conference on 4, 2129-2134, 1997
  4. A. Hvvarinen, J. Karhnen, and, E. Oja, Independent Component Anelvsis, (John Wiley & Sons, 2001)
  5. A. J. Bell and, T. J. Sejnowski, 'An informationmaximization approach to blind separation and blind deconvolution,' Neural Computation, 7 (6), 1129-1159, 1995 https://doi.org/10.1162/neco.1995.7.6.1129
  6. S. I. Amari, A. Cichocki and, H. H. Yang, 'A new learning algorithm for blind signal separation,' Advances in neural information Processinn systems 8. MIT Press, Cambridge, MA. 1995
  7. P. Smaragdis, 'Blind separation of convolved mixtures in the frequency domain,' Nerucomputing, 22, 21-34, 1998 https://doi.org/10.1016/S0925-2312(98)00047-2
  8. S. Araki, S. Makino, T. Nishikawa and, H. and Saruwatari, 'Fundamental limitation of frequency domain blind separation for convolutive of speech,' Proc. ICASSP 2001, 2737-2740, May. 2001
  9. N. Murata and, S. Ikeda, 'An approach to blind source separation based on temporal structure of speech signals' Neurocomputing 41. 1-24, 2001 https://doi.org/10.1016/S0925-2312(00)00345-3
  10. F. Asano, S.lkeda, M.Ogawa H.Asoh, and N. Kitawaki, 'A combined approach of array processing and independent component analysis for blind component analysis for blind separation of acoustic signals,' Proc. ICASSP 2001, 2729-2732, May 2001
  11. R. Lambert, Multichannel blind deconvolution: FIR matrix algebra and separation of multi path mixtures, Thesis, University of Southern California, Department of Electrical Engineering, May. 1996
  12. H. Saruwatari, T. Kawamura and, K. Shikano, 'Blind source separation for speech based on fastconvergence algorithm with ICA and beamforming,' Proc. Eurospeech 2001, 2603-2606, Sep. 2001
  13. W. Yan and, Z. Limimg, 'The effect of initial weight, learning rate and regularization on generalization performance and efficiency,' Proc. ICSP 2002, 1191-1194, Jun. 2002