Browse > Article

Recognition for Noisy Speech by a Nonstationary AR HMM with Gain Adaptation Under Unknown Noise  

이기용 (숭실대학교 정보통신전자공학부)
서창우 (숭실대학교 정보통신전자공학부)
이주헌 (동아방송대학 인터넷방송과)
Publication Information
The Journal of the Acoustical Society of Korea / v.21, no.1, 2002 , pp. 11-18 More about this Journal
Abstract
In this paper, a gain-adapted speech recognition method in noise is developed in the time domain. Noise is assumed to be colored. To cope with the notable nonstationary nature of speech signals such as fricative, glides, liquids, and transition region between phones, the nonstationary autoregressive (NAR) hidden Markov model (HMM) is used. The nonstationary AR process is represented by using polynomial functions with a linear combination of M known basis functions. When only noisy signals are available, the estimation problem of noise inevitably arises. By using multiple Kalman filters, the estimation of noise model and gain contour of speech is performed. Noise estimation of the proposed method can eliminate noise from noisy speech to get an enhanced speech signal. Compared to the conventional ARHMM with noise estimation, our proposed NAR-HMM with noise estimation improves the recognition performance about 2-3%.
Keywords
NAR-HMM; Multiple Kalman filters; EM algorithm; Speech enhancement; Speech recognition;
Citations & Related Records
연도 인용수 순위
  • Reference
1 Waveform-based Speech recogniton using hidden filter model: Parameter selection and sensitivity to power normalization /
[ H. Sheikhzadeh;L. Deng ] / IEEE Trans. on Speech and Audio Processing   DOI   ScienceOn
2 Filtering of Colored Noise for Speech Enhancement and Coding /
[ J. D. Gibson;B. Koo;S. D. Gray ] / IEEE Trans. on Signal Processing   DOI   ScienceOn
3 A Generalized hidden Markov model with state-conditioned trend functions of time for speech signal /
[ L. Deng ] / Signal Processing   DOI   ScienceOn
4 PMC for speech recognition in additive and convolutional noise /
[ M. J. F. Gales;S. J. Young ] / Technical Report CUED/F-INFENG/TR135
5 Gain adpted hidden Markov models for recognition of clean and noisy speech /
[ Y. Ephraim ] / IEEE Trans. Signal Processing   DOI   ScienceOn
6 Time-Dependent ARMA Modelling of Nonstationary Signals /
[ Y. Grenier ] / IEEE Trans. Acoust., Speech, Signal Processing   DOI
7 Eehancement of connected words in an extremely noisy environment /
[ Y. Cohen;A. Erell;Y. Bistritz ] / IEEE Trans. on Speech and Audio Processing
8 A nonstationary autoregressive HMM with gain adptation for speech recognition /
[ K. Y. Lee;J. Lee ] / Proc. ICSLP '98
9 A Maximization Technique in the statistical analysis of probabilstic functions of Markov chains /
[ L. E. Baum;T. Petrie;G. Soules;N. Weiss ] / Ann. Math. Stat.   DOI   ScienceOn
10 Subband Kalman filtering for speech enhancement /
[ W. Wu;P. Chen ] / IEEE Trans. on Circuits and Systems   DOI   ScienceOn
11 Mixture autoregressive hidden Markov models for speech signals /
[ B. Juang;L. R. Rabiner ] / IEEE Trans. Acoust.,Speech, Signal Processing   DOI
12 A stochastic model of speech incorporating hierarchical nonstationarity /
[ L. Deng ] / IEEE Trans. on Speech and Audio Processing   DOI   ScienceOn
13 Robustness in automatic speech recognition /
[ J. C. Junqua;J. P. Haton ] / Fundamentals and applications
14 On the application of the interacting multiple model algorithm for enhancing noisy speech /
[ J. B. Kim;K. Y. Lee;C. W. Lee ] / IEEE Trans. on Speech and Audio Processing
15 Speech recognition using HMM with polynomial regression functions as nonstationary states /
[ L. Deng;M. Aksmanovic;X. Sun;C. F. JeffWu ] / IEEE Trans. Speech and Audio Processing   DOI   ScienceOn
16 A nonstationary autoregressive HMM and its application to speech enhancement /
[ K. Y. Lee;J. Rheem ] / Proc. Eurospeech '97
17 Maximum likelihood from incomplete data via the EM Algorithm /
[ A. P. Dempster;N. M. Laird;D. B. Rubin ] / J. Royal Stat. Soc. B
18 A Markov model containing state-conditioned second-order nonstationary: Application to speech recognition /
[ L. Deng;C. Rathinavalu ] / Computer Speech and Language   DOI   ScienceOn