conference paper
On Optimal Smoothing in Minimum Statistics Based Noise Tracking
Aleksej
Chinaev
author
Reinhold
Haeb-Umbach
author 242
54
department
Noise tracking is an important component of speech enhancement algorithms. Of the many noise trackers proposed, Minimum Statistics (MS) is a particularly popular one due to its simple parameterization and at the same time excellent performance. In this paper we propose to further reduce the number of MS parameters by giving an alternative derivation of an optimal smoothing constant. At the same time the noise tracking performance is improved as is demonstrated by experiments employing speech degraded by various noise types and at different SNR values.
2015
eng
speech enhancementnoise trackingoptimal smoothing
Interspeech 2015
1785-1789
https://groups.uni-paderborn.de/nt/pubs/2015/ChHa15_Poster.pdf
Chinaev, Aleksej, and Reinhold Haeb-Umbach. “On Optimal Smoothing in Minimum Statistics Based Noise Tracking.” <i>Interspeech 2015</i>, 2015, pp. 1785–89.
Chinaev A, Haeb-Umbach R. On Optimal Smoothing in Minimum Statistics Based Noise Tracking. In: <i>Interspeech 2015</i>. ; 2015:1785-1789.
Chinaev, A., & Haeb-Umbach, R. (2015). On Optimal Smoothing in Minimum Statistics Based Noise Tracking. In <i>Interspeech 2015</i> (pp. 1785–1789).
A. Chinaev, R. Haeb-Umbach, in: Interspeech 2015, 2015, pp. 1785–1789.
A. Chinaev and R. Haeb-Umbach, “On Optimal Smoothing in Minimum Statistics Based Noise Tracking,” in <i>Interspeech 2015</i>, 2015, pp. 1785–1789.
Chinaev, Aleksej, and Reinhold Haeb-Umbach. “On Optimal Smoothing in Minimum Statistics Based Noise Tracking.” In <i>Interspeech 2015</i>, 1785–89, 2015.
@inproceedings{Chinaev_Haeb-Umbach_2015, title={On Optimal Smoothing in Minimum Statistics Based Noise Tracking}, booktitle={Interspeech 2015}, author={Chinaev, Aleksej and Haeb-Umbach, Reinhold}, year={2015}, pages={1785–1789} }
117392019-07-12T05:27:19Z2022-01-06T06:51:08Z