All Issue

2015 Vol.34, Issue 3 Preview Page
Speech Enhancement Using Nonnegative Matrix Factorization with Temporal Continuity

시간 연속성을 갖는 비음수 행렬 분해를 이용한 음질 개선

Seung-Hyon Nam1*
남 승현1*
1Department of Electronic Engineering,Paichai University
1배재대학교 전자공학과
*교신저자. *Corresponding Author. 
31 May 2015. pp. 240-246
PDF XML Citation
Abstract

In this paper, speech enhancement using nonnegative matrix factorization with temporal continuity has been addressed. Speech and noise signals are modeled as Possion distributions, and basis vectors and gain vectors of NMF are modeled as Gamma distributions. Temporal continuity of the gain vector is known to be critical to the quality of enhanced speech signals. In this paper, temporal continiuty is implemented by adopting Gamma-Markov chain priors for noise gain vectors during the separation phase. Simulation results show that the Gamma-Markov chain models temporal continuity of noise signals and track changes in noise effectively.

Keywords
Speech enhancement
Nonnegative matrix factorization
Variational Bayesian inference
Gamma-Makov chain
Temporal continuity

본 논문은 시간 연속성을 갖는 비음수 행렬 분해(Nonnegative Matrix Factorization, NMF)를 이용하여 잡음에 열화된 음성 신호의 음질을 개선하는 문제를 다룬다. 음성과 잡음 신호는 포아송 분포로 모델되며, NMF의 기본 벡터와 이득 벡터는 감마 분포로 모델된다. 이득 벡터의 시간 연속성은 음질 개선에 중요한 영향을 미치는 것으로 알려져 있다. 본 논문에서 시간의 연속성은 이득 벡터를 감마-마르코프 연쇄(Gamma-Markov chain, GMC) 사전 분포로 모델함으로써 이루어진다. 실험 결과는 제안된 알고리즘이 잡음 신호의 시간 연속성을 효과적으로 모델하는 것을 보여준다.

키워드
음질 개선
비음수행렬분해
변분-베이즈 추정
감마 -마르코프 연쇄
시간 연속성
References
1
1.P. Loizou, Speech Enhancement: Theory and Practice, 2nd Ed., (CRC Press, Inc. Boca Raton, FL, 2013), pp. 1-6.
2
2.P. Smaragdis, C. Fevotte, G. J. Mysore, N. Mohammadiha, and M. Hoffman, “Static and dynamic source separation using nonnegative factorization,” IEEE Sigal Processing Magazine, 66-75 (2014).
3
3.D. D. Lee and H. S. Seung, “Learning the parts of objects by non-negative matrix factorization,” Nature, 401, 788-791  (1999).
4
4.A. T. Cemgil, “Bayesian inference in non-negative matrix factorisation models,” Computational Intelligence and Neuroscience, 2009, Article ID 785152, 1-17 (2009).
5
5.C. M. Bishop, Pattern Recognition and Machine Learning (Springer, NewYork, 2006), pp. 462-466.
6
6.N. Mohammadiha, P. Smaragdis, and A. Leijon, “Supervised and unsupervised speech enhancement using NMF,” IEEE Trans. Audio, Speech, Lang. Processing, 21, 2140-2151, (2013).
7
7.T. Virtanen, A.T. Cemgil, and S. Godsill, “Bayesian extension to non-negative matrix factorization for audio signal modeling,” IEEE Int. Conf. on Acousts, Speech and Signal Process. 1825-1828, (2008).
8
8.http://www.uni-oldenburg.de/en/mediphysics-acoustics/ sigproc/staff/nasser-mohammadiha/matlab-codes/
9
9.A. T. Cemgil, and O. Dikmen, “Conjugate Gamma Markov random fields for modeling nonstationary sources,” 7th Int. Conf. on Independent Component Analysis and Signal Separation, 697-705, (2007)
10
10.S. R. Quackenbush, T. P. Barnwell, and M. A. Clements, Objective Measures of Speech Quality (Prentice Hall, New Jersey, 1988), pp. 45.
11
11.E. Vincent, R. Grivonval, and C. Févotte, “Performance measurement in blind audio source separation.” IEEE Trans. Audio, Speech, and Language Process. 14, 1462-1469 (2006).
Information
  • Publisher :The Acoustical Society of Korea
  • Publisher(Ko) :한국음향학회
  • Journal Title :The Journal of the Acoustical Society of Korea
  • Journal Title(Ko) :한국음향학회지
  • Volume : 34
  • No :3
  • Pages :240-246
  • Received Date : 2015-01-27
  • Revised Date : 2015-02-25
  • Accepted Date : 2015-03-27
  • DOI :https://doi.org/10.7776/ASK.2015.34.3.240
Journal Informaiton The Journal of the Acoustical Society of Korea The Journal of the Acoustical Society of Korea
E-Book
  • scopus_JASK
  • ESCI_JASK
  • Google scholar_JASK
  • NRF
  • KOFST
  • KISTI Current Status
  • KISTI Cited-by
  • crosscheck
  • orcid
  • ccl
Journal Informaiton Journal Informaiton - close