Speech Enhancement Based on Adaptive Noise Power Estimation Using Spectral Difference
Summary :
In this paper, we propose a spectral difference approach for noise power estimation in speech enhancement. The noise power estimate is given by recursively averaging past spectral power values using a smoothing parameter based on the current observation. The smoothing parameter in time and frequency is adjusted by the spectral difference between consecutive frames that can efficiently characterize noise variation. Specifically, we propose an effective technique based on a sigmoid-type function in order to adaptively determine the smoothing parameter based on the spectral difference. Compared to a conventional method, the proposed noise estimate is computationally efficient and able to effectively follow noise changes under various noise conditions.
- Publication
- IEICE TRANSACTIONS on Fundamentals Vol.E94-A No.10 pp.2031-2034
- Publication Date
- 2011/10/01
- Publicized
- Online ISSN
- 1745-1337
- DOI
- 10.1587/transfun.E94.A.2031
- Type of Manuscript
- LETTER
- Category
- Speech and Hearing
Authors
Keyword
Latest Issue
Copyrights notice of machine-translated contents
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Cite this
Copy
Jae-Hun CHOI, Joon-Hyuk CHANG, Dong Kook KIM, Suhyun KIM, "Speech Enhancement Based on Adaptive Noise Power Estimation Using Spectral Difference" in IEICE TRANSACTIONS on Fundamentals,
vol. E94-A, no. 10, pp. 2031-2034, October 2011, doi: 10.1587/transfun.E94.A.2031.
Abstract: In this paper, we propose a spectral difference approach for noise power estimation in speech enhancement. The noise power estimate is given by recursively averaging past spectral power values using a smoothing parameter based on the current observation. The smoothing parameter in time and frequency is adjusted by the spectral difference between consecutive frames that can efficiently characterize noise variation. Specifically, we propose an effective technique based on a sigmoid-type function in order to adaptively determine the smoothing parameter based on the spectral difference. Compared to a conventional method, the proposed noise estimate is computationally efficient and able to effectively follow noise changes under various noise conditions.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/transfun.E94.A.2031/_p
Copy
@ARTICLE{e94-a_10_2031,
author={Jae-Hun CHOI, Joon-Hyuk CHANG, Dong Kook KIM, Suhyun KIM, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Speech Enhancement Based on Adaptive Noise Power Estimation Using Spectral Difference},
year={2011},
volume={E94-A},
number={10},
pages={2031-2034},
abstract={In this paper, we propose a spectral difference approach for noise power estimation in speech enhancement. The noise power estimate is given by recursively averaging past spectral power values using a smoothing parameter based on the current observation. The smoothing parameter in time and frequency is adjusted by the spectral difference between consecutive frames that can efficiently characterize noise variation. Specifically, we propose an effective technique based on a sigmoid-type function in order to adaptively determine the smoothing parameter based on the spectral difference. Compared to a conventional method, the proposed noise estimate is computationally efficient and able to effectively follow noise changes under various noise conditions.},
keywords={},
doi={10.1587/transfun.E94.A.2031},
ISSN={1745-1337},
month={October},}
Copy
TY - JOUR
TI - Speech Enhancement Based on Adaptive Noise Power Estimation Using Spectral Difference
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 2031
EP - 2034
AU - Jae-Hun CHOI
AU - Joon-Hyuk CHANG
AU - Dong Kook KIM
AU - Suhyun KIM
PY - 2011
DO - 10.1587/transfun.E94.A.2031
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E94-A
IS - 10
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - October 2011
AB - In this paper, we propose a spectral difference approach for noise power estimation in speech enhancement. The noise power estimate is given by recursively averaging past spectral power values using a smoothing parameter based on the current observation. The smoothing parameter in time and frequency is adjusted by the spectral difference between consecutive frames that can efficiently characterize noise variation. Specifically, we propose an effective technique based on a sigmoid-type function in order to adaptively determine the smoothing parameter based on the spectral difference. Compared to a conventional method, the proposed noise estimate is computationally efficient and able to effectively follow noise changes under various noise conditions.
ER -
English
