Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech
V. Leutnant, A. Krueger, R. Haeb-Umbach, Speech Communication; 10. ITG Symposium; Proceedings Of (2012) 1–4.
Download (ext.)
Journal Article
| English
Author
Leutnant, Volker;
Krueger, Alexander;
Haeb-Umbach, ReinholdLibreCat
Abstract
In this contribution, a new observation model for the joint compensation of reverberation and noise in the logarithmic mel power spectral density domain will be considered. The proposed observation model relates the noisy reverberant feature to the underlying sequence of clean speech features and the feature of the noise. Nevertheless, due to the complex interaction of these variables in the target domain, the observationmodel cannot be applied to Bayesian feature enhancement directly, calling for approximations that eventually render the observation model useful. The performance of the approximated observation model will highly depend on the capability of modeling the difference between the model and the noisy reverberant observation. A detailed analysis of this observation error will be provided in this work. Among others, it will point out the need to account for the instantaneous ratio of the reverberant speech power and the noise power. Index Terms: Bayesian feature enhancement, observation model for noisy reverberant speech
Publishing Year
Journal Title
Speech Communication; 10. ITG Symposium; Proceedings of
Page
1-4
LibreCat-ID
Cite this
Leutnant V, Krueger A, Haeb-Umbach R. Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech. Speech Communication; 10 ITG Symposium; Proceedings of. 2012:1-4.
Leutnant, V., Krueger, A., & Haeb-Umbach, R. (2012). Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech. Speech Communication; 10. ITG Symposium; Proceedings Of, 1–4.
@article{Leutnant_Krueger_Haeb-Umbach_2012, title={Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech}, journal={Speech Communication; 10. ITG Symposium; Proceedings of}, author={Leutnant, Volker and Krueger, Alexander and Haeb-Umbach, Reinhold}, year={2012}, pages={1–4} }
Leutnant, Volker, Alexander Krueger, and Reinhold Haeb-Umbach. “Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech.” Speech Communication; 10. ITG Symposium; Proceedings Of, 2012, 1–4.
V. Leutnant, A. Krueger, and R. Haeb-Umbach, “Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech,” Speech Communication; 10. ITG Symposium; Proceedings of, pp. 1–4, 2012.
Leutnant, Volker, et al. “Investigations Into a Statistical Observation Model for Logarithmic Mel Power Spectral Density Features of Noisy Reverberant Speech.” Speech Communication; 10. ITG Symposium; Proceedings Of, 2012, pp. 1–4.
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