Advanced

Rayleigh mixture model-based hidden Markov modeling and estimation of noise in noisy speech signals

Sorensen, Karsten Vandborg and Andersen, Sören Vang LU (2007) In IEEE Transactions on Audio, Speech, and Language Processing 15(3). p.901-917
Abstract
In this paper, we propose a new statistical model for noise periodogram modeling and estimation. The proposed model is a hidden Markov model (HMM) with a Rayleigh mixture model (RMM) in each state. For this new model, we derive an expectation-maxinuzation (EM) training algorithm and a minimum meansquare error (NIMSE) noise periodogram estimator. It is shown that when compared to the Gaussian mixture model (GMM)-based HMM, the RMM-based HMM has less computationally complex EM iterations and gives a better fit of the noise periodograms when the mixture models has a low number of components. Furthermore, we propose a specialization of the proposed model, which is shown to provide better MMSE noise periodograrn estimates than any other of the... (More)
In this paper, we propose a new statistical model for noise periodogram modeling and estimation. The proposed model is a hidden Markov model (HMM) with a Rayleigh mixture model (RMM) in each state. For this new model, we derive an expectation-maxinuzation (EM) training algorithm and a minimum meansquare error (NIMSE) noise periodogram estimator. It is shown that when compared to the Gaussian mixture model (GMM)-based HMM, the RMM-based HMM has less computationally complex EM iterations and gives a better fit of the noise periodograms when the mixture models has a low number of components. Furthermore, we propose a specialization of the proposed model, which is shown to provide better MMSE noise periodograrn estimates than any other of the tested HMA4 initializations for cyclo-stationary noise types. (Less)
Please use this url to cite or link to this publication:
author
publishing date
type
Contribution to journal
publication status
published
subject
keywords
hidden Markov model (HMM), probability density function, Rayleigh, mixture model (RMM), speech enhancement
in
IEEE Transactions on Audio, Speech, and Language Processing
volume
15
issue
3
pages
901 - 917
publisher
IEEE--Institute of Electrical and Electronics Engineers Inc.
external identifiers
  • wos:000244318600015
  • scopus:64149128862
ISSN
1558-7924
DOI
10.1109/TASL.2006.885240
language
English
LU publication?
no
id
ae9d69e3-26f0-4366-a8e2-eff5fe7f65e2 (old id 4092498)
date added to LUP
2013-10-17 10:56:11
date last changed
2017-01-01 04:24:07
@article{ae9d69e3-26f0-4366-a8e2-eff5fe7f65e2,
  abstract     = {In this paper, we propose a new statistical model for noise periodogram modeling and estimation. The proposed model is a hidden Markov model (HMM) with a Rayleigh mixture model (RMM) in each state. For this new model, we derive an expectation-maxinuzation (EM) training algorithm and a minimum meansquare error (NIMSE) noise periodogram estimator. It is shown that when compared to the Gaussian mixture model (GMM)-based HMM, the RMM-based HMM has less computationally complex EM iterations and gives a better fit of the noise periodograms when the mixture models has a low number of components. Furthermore, we propose a specialization of the proposed model, which is shown to provide better MMSE noise periodograrn estimates than any other of the tested HMA4 initializations for cyclo-stationary noise types.},
  author       = {Sorensen, Karsten Vandborg and Andersen, Sören Vang},
  issn         = {1558-7924},
  keyword      = {hidden Markov model (HMM),probability density function,Rayleigh,mixture model (RMM),speech enhancement},
  language     = {eng},
  number       = {3},
  pages        = {901--917},
  publisher    = {IEEE--Institute of Electrical and Electronics Engineers Inc.},
  series       = {IEEE Transactions on Audio, Speech, and Language Processing},
  title        = {Rayleigh mixture model-based hidden Markov modeling and estimation of noise in noisy speech signals},
  url          = {http://dx.doi.org/10.1109/TASL.2006.885240},
  volume       = {15},
  year         = {2007},
}