Title :
Robust noise reduction for speech and audio signals
Author :
Godsill, Simon J. ; Rayner, Peter J W
Author_Institution :
Dept. of Eng., Cambridge Univ., UK
Abstract :
Statistical model-based methods are presented for the reconstruction of autocorrelated signals in impulsive plus continuous noise environments. Signals are modelled as autoregressive and noise sources as discrete and continuous mixtures of Gaussians, allowing for robustness in highly impulsive and non-Gaussian environments. Markov Chain Monte Carlo methods are used for reconstruction of the corrupted waveforms within a Bayesian probabilistic framework and results are presented for contaminated voice and audio signals
Keywords :
Bayes methods; Gaussian noise; Markov processes; Monte Carlo methods; acoustic signal detection; acoustic signal processing; audio signals; probability; signal reconstruction; speech enhancement; speech processing; Bayesian impulse detection; Bayesian probabilistic framework; Markov chain Monte Carlo methods; audio signals; autocorrelated signal reconstruction; autoregressive sources; continuous Gaussian mixtures; continuous noise environment; corrupted waveforms; discrete Gaussian mixtures; impulsive noise environment; noise sources; nonGaussian environment; robust noise reduction; speech signals; statistical model based methods; Background noise; Bayesian methods; Electromagnetic interference; Gaussian noise; Image reconstruction; Noise reduction; Noise robustness; Signal processing; Speech enhancement; Working environment noise;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1996. ICASSP-96. Conference Proceedings., 1996 IEEE International Conference on
Conference_Location :
Atlanta, GA
Print_ISBN :
0-7803-3192-3
DOI :
10.1109/ICASSP.1996.543198
