MCMC methods for restoration of nonlinearly distorted autoregressive signals

We approach the problem of restoring distorted autoregressive (AR) signals by using a cascade model, in which the observed signal is modelled as the output of a nonlinear AR process (NAR) excited by the linear AR signal we are attempting to recover. The Volterra expansion of the NAR model has a very large number of possible terms even when truncated at fairly small maximum orders and lags. We address the problem of subset selection and uncertainty in the nonlinear stage and model length uncertainty in the linear stage through a hierarchical Bayesian approach, using reversible jump Markov chain Monte Carlo (MCMC) and Gibbs sampling. We demonstrate the method using synthetic AR data, and extend the approach to process a long distorted audio time series, for which the source model cannot be considered to be stationary.

[1]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[2]  Malcolm J. Hawksford,et al.  Identifica-tion of Discrete Volterra Series Using Maximum Length Sequences , 1996 .

[3]  Lynn Kuo,et al.  Bayesian semiparametric inference for the accelerated failure‐time model , 1997 .

[4]  Guoping Liu,et al.  On-line identification of nonlinear systems using volterra polynomial basis function neural networks , 1997, 1997 European Control Conference (ECC).

[5]  H. Akaike.  A new look at the statistical model identification , 1974 .

[6]  Simon J. Godsill,et al.  Statistical reconstruction and analysis of autoregressive signals in impulsive noise using the Gibbs sampler , 1998, IEEE Trans. Speech Audio Process..

[7]  Jonathan D. Cryer,et al.  Time Series Analysis , 1986 .

[8]  Simon J. Godsill,et al.  Bayesian model selection for linear and non-linear time series using the Gibbs sampler , 1998 .

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

[10]  L. Tierney.  Markov Chains for Exploring Posterior Distributions , 1994 .

[11]  Stanley A. White.  Restoration of Nonlinearly Distorted Audio by Histogram Equalization , 1981 .

[12]  Sylvia Richardson,et al.  Stochastic search variable selection , 1995 .

[13]  W. K. Hastings,et al.  Monte Carlo Sampling Methods Using Markov Chains and Their Applications , 1970 .

[14]  Malcolm J. Hawksford,et al.  Practical modelling of nonlinear audio sys-tems using the Volterra series , 1996 .

[15]  Wj Fitzgerald,et al.  Interpolation of missing samples for audio restoration , 1994 .

[16]  Simon J. Godsill,et al.  Robust treatment of impulsive noise in speech and audio signals , 1996 .

[17]  S. V. Vaseghi,et al.  Restoration of Old Gramophone Recordings , 1992 .

[18]  H. Tong.  Non-linear time series. A dynamical system approach , 1990 .

[19]  E. George,et al.  Journal of the American Statistical Association is currently published by American Statistical Association. , 2007 .

[20]  G. Roberts,et al.  Updating Schemes, Correlation Structure, Blocking and Parameterization for the Gibbs Sampler , 1997 .

[21]  O. E. Herrmann,et al.  Second order Volterra inverses for compensation of loudspeaker nonlinearity , 1995, Proceedings of 1995 Workshop on Applications of Signal Processing to Audio and Accoustics.

[22]  Peter J. W. Rayner,et al.  Unsupervised time series classification , 1995, Signal Process..

[23]  Wolfgang Klippel.  Adaptive Adjustment of Nonlinear Filters Used for Loudspeaker Linearization , 1998 .

[24]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1971 .

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

[26]  Simon J. Godsill,et al.  Bayesian model selection for time series using Markov chain Monte Carlo , 1997, 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[27]  Simon J. Godsill,et al.  Robust noise reduction for speech and audio signals , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[28]  Simon J. Godsill,et al.  A reversible jump sampler for autoregressive time series , 1997, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[29]  V. Haggan,et al.  ON THE SELECTION OF SUBSET AUTOREGRESSIVE TIME SERIES MODELS , 1984 .

[30]  浜田 晴夫,et al.  1989 IEEE ASSP Workshop on Applications of Signal Processing to Audio and Acoustics , 1990 .

[31]  Simon J. Godsill,et al.  Robust noise modelling with application to audio restoration , 1995, Proceedings of 1995 Workshop on Applications of Signal Processing to Audio and Accoustics.

[32]  Simon J. Godsill,et al.  Digital audio restoration , 1998 .

[33]  Sheng Chen,et al.  Identification of MIMO non-linear systems using a forward-regression orthogonal estimator , 1989 .

[34]  P. Young,et al.  Time series analysis, forecasting and control , 1972, IEEE Transactions on Automatic Control.

[35]  Douglas Preis,et al.  Restoration of nonlinearly distorted magnetic recordings , 1983, ICASSP.

[36]  J. Geweke,et al.  Variable selection and model comparison in regression , 1994 .

[37]  R. de Figueiredo.  The Volterra and Wiener theories of nonlinear systems , 1982, Proceedings of the IEEE.

[38]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[39]  Simon J. Godsill,et al.  A Bayesian approach to the restoration of degraded audio signals , 1995, IEEE Trans. Speech Audio Process..

[40]  James D. Hamilton.  Time Series Analysis , 1994 .

[41]  Adrian F. M. Smith,et al.  Bayesian Statistics 5. , 1998 .

[42]  Sheng Chen,et al.  Modelling and analysis of non-linear time series , 1989 .

[43]  P. T. Troughton.  Bayesian restoration of quantised audio signals using a sinusoidal model with autoregressive residuals , 1999, Proceedings of the 1999 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics. WASPAA'99 (Cat. No.99TH8452).

[44]  John H. L. Hansen,et al.  Discrete-Time Processing of Speech Signals , 1993 .

[45]  Jun S. Liu,et al.  Covariance Structure and Convergence Rate of the Gibbs Sampler with Various Scans , 1995 .

[46]  Keith Worden,et al.  FREQUENCY DOMAIN ANALYSIS OF NARX NEURAL NETWORKS , 1998 .