Particle methods for Bayesian modeling and enhancement of speech signals

This paper applies time-varying autoregressive (TVAR) models with stochastically evolving parameters to the problem of speech modeling and enhancement. The stochastic evolution models for the TVAR parameters are Markovian diffusion processes. The main aim of the paper is to perform on-line estimation of the clean speech and model parameters and to determine the adequacy of the chosen statistical models. Efficient particle methods are developed to solve the optimal filtering and fixed-lag smoothing problems. The algorithms combine sequential importance sampling (SIS), a selection step and Markov chain Monte Carlo (MCMC) methods. They employ several variance reduction strategies to make the best use of the statistical structure of the model. It is also shown how model adequacy may be determined by combining the particle filter with frequentist methods. The modeling and enhancement performance of the models and estimation algorithms are evaluated in simulation studies on both synthetic and real speech data sets.

[1]  A. Willsky,et al.  Time-varying parametric modeling of speech☆ , 1983 .

[2]  Amir Dembo,et al.  Maximum a posteriori estimation of time-varying ARMA processes from noisy observations , 1988, IEEE Trans. Acoust. Speech Signal Process..

[3]  Peter J. W. Rayner,et al.  Digital Audio Restoration: A Statistical Model Based Approach , 1998 .

[4]  R. Kohn,et al.  Bayesian estimation of an autoregressive model using Markov chain Monte Carlo , 1996 .

[5]  Nando de Freitas,et al.  Sequential Monte Carlo Methods in Practice , 2001, Statistics for Engineering and Information Science.

[6]  Jun S. Liu,et al.  Sequential Imputations and Bayesian Missing Data Problems , 1994 .

[7]  B. Anderson,et al.  Optimal Filtering , 1979, IEEE Transactions on Systems, Man, and Cybernetics.

[8]  Yves Grenier,et al.  Time-dependent ARMA modeling of nonstationary signals , 1983 .

[9]  Andrew Harvey,et al.  Forecasting, Structural Time Series Models and the Kalman Filter , 1990 .

[10]  L. A. Liporace.  Linear estimation of nonstationary signals. , 1975, The Journal of the Acoustical Society of America.

[11]  J. Makhoul,et al.  Linear prediction: A tutorial review , 1975, Proceedings of the IEEE.

[12]  G. Box,et al.  On a measure of lack of fit in time series models , 1978 .

[13]  N. Gordon,et al.  Novel approach to nonlinear/non-Gaussian Bayesian state estimation , 1993 .

[14]  A.V. Oppenheim,et al.  Enhancement and bandwidth compression of noisy speech , 1979, Proceedings of the IEEE.

[15]  Jun S. Liu,et al.  Sequential Monte Carlo methods for dynamic systems , 1997 .

[16]  N. Shephard,et al.  Stochastic Volatility: Likelihood Inference And Comparison With Arch Models , 1996 .

[17]  R. Kohn,et al.  Diagnostics for Time Series Analysis , 1999 .

[18]  H. Sorenson,et al.  Nonlinear Bayesian estimation using Gaussian sum approximations , 1972 .

[19]  Ehud Weinstein,et al.  Iterative and sequential Kalman filter-based speech enhancement algorithms , 1998, IEEE Trans. Speech Audio Process..

[20]  G. Kitagawa.  Monte Carlo Filter and Smoother for Non-Gaussian Nonlinear State Space Models , 1996 .

[21]  D. Mayne,et al.  Monte Carlo techniques to estimate the conditional expectation in multi-stage non-linear filtering† , 1969 .

[22]  M. Rosenblatt.  Remarks on a Multivariate Transformation , 1952 .

[23]  R. Bucy,et al.  Digital synthesis of non-linear filters , 1971 .

[24]  Simon J. Godsill,et al.  Non-stationary Bayesian modelling and enhancement of speech signals , 1999 .

[25]  Masahito Yamada,et al.  Structural Time Series Models and the Kalman Filter , 1989 .

[26]  Hoon Kim,et al.  Monte Carlo Statistical Methods , 2000, Technometrics.

[27]  Alan Oppenheim,et al.  Time-varying parametric modeling of speech , 1977, 1977 IEEE Conference on Decision and Control including the 16th Symposium on Adaptive Processes and A Special Symposium on Fuzzy Set Theory and Applications.

[28]  Pierre Del Moral,et al.  Discrete Filtering Using Branching and Interacting Particle Systems , 1998 .

[29]  L. Shenton,et al.  Omnibus test contours for departures from normality based on √b1 and b2 , 1975 .

[30]  Jun S. Liu,et al.  Blind Deconvolution via Sequential Imputations , 1995 .

[31]  Simon J. Godsill,et al.  On sequential Monte Carlo sampling methods for Bayesian filtering , 2000, Stat. Comput..