Theoretical analysis of adaptive noise reduction algorithms for hearing aids

Author

Maj, Jean Baptiste ; Moonen, Marc ; Wouters, Jan

Author_Institution

ESAT-SISTA, KU Leuven, Leuven, Belgium

fYear

2002

fDate

3-6 Sept. 2002

Firstpage

1

Lastpage

4

Abstract

This paper presents the theoretical analysis of 2 different adaptive noise reduction algorithms for twin-microphone hearing aids. A first noise reduction algorithm is based on a beamformer technique [1] and a second is based on optimal filtering and singular value decomposition (SVD) [2]. On the one hand, it is shown that the SVD-based technique gives a robust solution against unmatched microphone characteristics. On the other hand, the beamformer technique has a better robustness against voice activity detector (VAD) errors.

Keywords

array signal processing; hearing aids; medical signal processing; microphones; signal denoising; singular value decomposition; speech recognition; SVD-based technique; VAD; adaptive noise reduction algorithms; beamformer technique; optimal filtering; singular value decomposition; twin-microphone hearing aids; unmatched microphone characteristics; voice activity detector errors; Algorithm design and analysis; Filtering algorithms; Maximum likelihood detection; Noise; Noise reduction; Nonlinear filters; Software algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Conference, 2002 11th European

Conference_Location

Toulouse

ISSN

2219-5491

Type

conf

Filename

7071961