DocumentCode :
3118330
Title :
Single-channel speech enhancement: Using recurrent neuro-fuzzy voice activity detector and spectral subtraction algorithms
Author :
Chuang, Fang-Chen ; Wang, Jeen-Shing ; Wu, Li-Ying
Author_Institution :
Dept. of Electr. Eng., Nat. Cheng Kung Univ., Tainan
fYear :
2008
fDate :
12-15 Oct. 2008
Firstpage :
3057
Lastpage :
3061
Abstract :
This paper investigates the effectiveness of a single-channel speech enhancement system that contains spectral subtraction and voice activity detection algorithm for noise elimination. We first extract features from a noisy signal and use these features as the inputs of a recurrent neuro-fuzzy network for detecting the voice activities of the signal. Based on the detection, we describe the characteristics of the background noise of the speech segments by a minimum frequency energy (MFE) parameter and then apply spectral subtraction algorithms with the parameter to eliminate the noise. Our simulation results show that the proposed enhancement system with a nonlinear spectral subtraction algorithm has superior performance.
Keywords :
feature extraction; fuzzy neural nets; recurrent neural nets; signal detection; spectral analysis; speech enhancement; feature extraction; minimum frequency energy parameter; noise elimination; nonlinear spectral subtraction algorithm; recurrent neuro-fuzzy voice activity detector; single-channel speech enhancement; Background noise; Detection algorithms; Detectors; Feature extraction; Filter bank; Frequency; Fuzzy neural networks; Speech enhancement; Time domain analysis; Working environment noise; Speech enhancement; recurrent networks; spectral; subtraction; voice detection;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Systems, Man and Cybernetics, 2008. SMC 2008. IEEE International Conference on
Conference_Location :
Singapore
ISSN :
1062-922X
Print_ISBN :
978-1-4244-2383-5
Electronic_ISBN :
1062-922X
Type :
conf
DOI :
10.1109/ICSMC.2008.4811764
Filename :
4811764
Link To Document :
بازگشت