DocumentCode :
3391244
Title :
Bayesian restoration of quantised audio signals using a sinusoidal model with autoregressive residuals
Author :
Troughton, Paul Z.
Author_Institution :
Dept. of Eng., Cambridge Univ., UK
fYear :
1999
fDate :
1999
Firstpage :
159
Lastpage :
162
Abstract :
In digital audio systems, the amplitude of the signal is quantised with finite resolution. This is a nonlinear process which introduces distortion. We develop a Bayesian, model-based approach to reducing quantisation distortion when moving an audio signal to a higher resolution medium. The signal is modelled as a sum of sinusoids and an autoregressive (AR) process of unknown order. Estimation is performed using Markov chain Monte Carlo (MCMC) methods. Selection of the correct AR model order and number of sinusoids is found to be crucial to avoiding artefacts; both of these operations are incorporated into the MCMC structure. An approximate, but fast converging, MCMC method is developed for selecting the sinusoids, and moves between AR models of different orders are made using reversible-jump methods
Keywords :
Bayes methods; Markov processes; Monte Carlo methods; approximation theory; audio signal processing; audio systems; autoregressive processes; convergence of numerical methods; digital communication; quantisation (signal); signal restoration; AR model order selection; AR process; Bayesian model-based approach; Bayesian restoration; Markov chain Monte Carlo MCMC methods; approximate fast converging MCMC method; artefacts avoidance; autoregressive process; autoregressive residuals; digital audio systems; distortion; model noise reduction; nonlinear process; quantisation distortion; quantised audio signals; reversible-jump methods; signal amplitude; sinusoidal model; Audio systems; Bayesian methods; Digital signal processing; Distortion; Monte Carlo methods; Quantization; Signal processing; Signal resolution; Signal restoration; Signal sampling;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Applications of Signal Processing to Audio and Acoustics, 1999 IEEE Workshop on
Conference_Location :
New Paltz, NY
Print_ISBN :
0-7803-5612-8
Type :
conf
DOI :
10.1109/ASPAA.1999.810874
Filename :
810874
Link To Document :
بازگشت