Modeling the dynamics of speech and noise for speech feature enhancement in ASR
S. Windmann, R. Haeb-Umbach, in: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008), 2008, pp. 4409–4412.
Download (ext.)
Conference Paper
| English
Author
Windmann, Stefan;
Haeb-Umbach, ReinholdLibreCat
Abstract
In this paper a switching linear dynamical model (SLDM) approach for speech feature enhancement is improved by employing more accurate models for the dynamics of speech and noise. The model of the clean speech feature trajectory is improved by augmenting the state vector to capture information derived from the delta features. Further a hidden noise state variable is introduced to obtain a more elaborated model for the noise dynamics. Approximate Bayesian inference in the SLDM is carried out by a bank of extended Kalman filters, whose outputs are combined according to the a posteriori probability of the individual state models. Experimental results on the AURORA2 database show improved recognition accuracy.
Keywords
a posteriori probability;
AURORA2 database;
Bayesian inference;
Bayes methods;
channel bank filters;
extended Kalman filter banks;
hidden noise state variable;
Kalman filters;
noise dynamics;
speech enhancement;
speech feature enhancement;
speech feature trajectory;
switching linear dynamical model approach
Publishing Year
Proceedings Title
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008)
Page
4409-4412
LibreCat-ID
Cite this
Windmann S, Haeb-Umbach R. Modeling the dynamics of speech and noise for speech feature enhancement in ASR. In: IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008). ; 2008:4409-4412. doi:10.1109/ICASSP.2008.4518633
Windmann, S., & Haeb-Umbach, R. (2008). Modeling the dynamics of speech and noise for speech feature enhancement in ASR. In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008) (pp. 4409–4412). https://doi.org/10.1109/ICASSP.2008.4518633
@inproceedings{Windmann_Haeb-Umbach_2008, title={Modeling the dynamics of speech and noise for speech feature enhancement in ASR}, DOI={10.1109/ICASSP.2008.4518633}, booktitle={IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008)}, author={Windmann, Stefan and Haeb-Umbach, Reinhold}, year={2008}, pages={4409–4412} }
Windmann, Stefan, and Reinhold Haeb-Umbach. “Modeling the Dynamics of Speech and Noise for Speech Feature Enhancement in ASR.” In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008), 4409–12, 2008. https://doi.org/10.1109/ICASSP.2008.4518633.
S. Windmann and R. Haeb-Umbach, “Modeling the dynamics of speech and noise for speech feature enhancement in ASR,” in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008), 2008, pp. 4409–4412.
Windmann, Stefan, and Reinhold Haeb-Umbach. “Modeling the Dynamics of Speech and Noise for Speech Feature Enhancement in ASR.” IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2008), 2008, pp. 4409–12, doi:10.1109/ICASSP.2008.4518633.
All files available under the following license(s):
Copyright Statement:
This Item is protected by copyright and/or related rights. [...]
Link(s) to Main File(s)
Access Level
Closed Access
Google Scholar