The Journal of the Acoustical Society of Korea (한국음향학회지)

The Acoustical Society of Korea (한국음향학회)
권호 표지

Implementation of a Robust Speaker Recognition System in Noisy Environment Using AR HMM with Duration-term

지속시간항을 갖는 AR HMM을 이용한 잡음환경에서의 강인 화자인식 시스템 구현

  • 이기용 (숭실대학교 정보통신 전자공학부) ;
  • 임재열 (한국기술교육대학교 정보기술공학부)
  • Published : 2001.08.01
Download PDF
⟨ Previous Next ⟩

Abstract

Though speaker recognition based on conventional AR HMM shows good performance, its lack of modeling the environmental noise makes its performance degraded in case of practical noisy environment. In this paper, a robust speaker recognition system based on AR HMM is proposed, where noise is considered in the observation signal model for practical noisy environment and duration-term is considered to increase performance. Experimental results, using the digits database from 100 speakers (77 males and 23 females) under white noise and car noise, show improved performance.

기존의 AR HMM(auroreg ressive hidden morkov model)에 의한 화자인식 방법은 그 성능이 우수하나, 잡음에 대한 것이 고려되지 않아 실제 환경에 적용시 성능저하가 문제가 된다. 본 논문에서는 실제 환경에 맞추기 위하여 관측 신호 모델에서 잡음을 고려하고, 화자인식 성능을 개선하고자 지속시간항 (duration-term)을 포함하는 AR HMM을 이용하여 잡음환경에서의 강인한 화자인식 시스템을 제안한다. 100명의 화자 (남자 77명, 여자 23명)가 2주에 걸쳐 6번 발성한 숫자음 데이터베이스을 가지고, 백색잡음 및 자동차 잡음하에서 실험한 결과, 제안된 방법으로 성능이 향상됨을 확인하였다.

Keywords

References

  1. Automatic Speech & Speaker Recognition N. R. Dixon;T. B. Martin
  2. J. Acoust. Soc. Amer. v.55 Effectiveness of linear predictive characteristics of the Speech wave for automatic speaker identification and verification B. S. Atal
  3. Proc. ICASSP-90 Sub-word unit talker verification using HMM A. E. Rosenberg(et al.)
  4. IEEE Trans. ASSP v.39 On the application of mixture AR HMM to text independent speaker recognition N. Tishby
  5. IEE Trans. SAP v.2 Comparison of text-independent speaker recognition method using VQ-distortion and HMM T. Matsui;S. Furui
  6. EURASIP Signal Prcdssing v.27 A generalized hidden Markov model with state-conditioned trend functions of time for the speech signal L. Deng
  7. IEEE Signal Processing Letters v.3 Efficient recursive estimation for speech enhancement in colored noise K. Y. Lee;K. Shirai
  8. IEEE APCCAS-96 Recursive estimation based on the trend Hidden Markov Model in speech enhancement K. Y. Lee;J. Y. RHEEM;K. Shirai
  9. IEEE Trans., ASSP v.36 no.4 Time-dependent ARMA modeling of nonstationary signals Y. Grenier
  10. IEEE Trans., SP Robust parametric modeling of durations in Hidden Markov Models D. Burshtein
  11. ECSA EUROSPEECH-95 Speaker recognition using HMM composition in noisy environments T. Matsui;T. Kanno;S. Furui
  12. IEEE Trans., AC State estimation for systems with sojourn-Time-Dependent Markov model switching L. Campo;P. Mookerjee;Y. Bar-Shalom
  13. Signal processing Nonstationary hidden Markov model B. K. Sin;J. H. Kim
  14. IEEE Trans., Speech and Audio proc. v.4 no.1 Speaker verification using normalized log-likelihood score C. S. Liu;H. C. Wang;C. H. Lee
  15. IEEE Trans., Speech and Audio proc. v.3 no.1 Robust text-independent speaker identification using Gaussian mixture speaker models D. A. Reynolds;R. C. Rose
  16. IEEE Signal processing letters v.2 no.3 Large population speaker identification using clean and telephone speech D. A. Reynolds
  17. IEEE Trans., Speech and Audio Proc. v.2 no.4 Experimental evaluation of features for robust speaker identification D. A. Reynolds
  18. Fundamental of Speech Recognition L. R. Rabiner;B. H. Junag
  19. IEEE Trans. Speech and Audio Proc. v.3 no.3 On the Complexity of explicit Duration HMM's C. Mitchell;M. Harper;L. Jamieson
  20. Signal Processing v.46 An EM-based approach for parameter enhancement with an application to speech signals B.-G. Lee;K. Y. Lee;S. Ann
  21. IEEE Signal Processing Letters v.4 no.10 Adaptive filtering for speech enhancement in colored noise K. Y. Lee;B-G. Lee;S. Ann