Neuro-fuzzy filtering techniques for automatic speech recognition enhancement

Author

Poluzzi, Rinaldo ; Arnone, L. ; Savi, Alberto ; Brescianini, Michele

Author_Institution

Centro Direzionale Colleoni, ST Microelectron., Milano, Italy

fYear

2003

fDate

4-6 Sept. 2003

Firstpage

255

Lastpage

258

Abstract

In order to enhance the percentage of recognised words for ASR (automatic speech recognition) systems, an adaptive neuro-fuzzy filtering technique in the time domain is proposed. Architecture of neural networks is shown having the purpose of exploiting spatial information, with the advantage of little computational load with respect to other classical beamforming techniques. Several results are described for complex acoustic scenarios with VOX ASR system by DIBE (University of Genova).

Keywords

fuzzy neural nets; neural net architecture; nonlinear filters; signal denoising; speech enhancement; speech recognition; VOX ASR system; automatic speech recognition enhancement; beamforming techniques; computational load; neural networks; neuro-fuzzy filtering techniques; neuro-fuzzy network; nonlinear filtering; spatial information; Adaptive filters; Array signal processing; Automatic speech recognition; Computer architecture; Filtering; Fuzzy neural networks; Noise reduction; Nonlinear filters; Speech enhancement; Speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Signal Processing, 2003 IEEE International Symposium on

Print_ISBN

0-7803-7864-4

Type

conf

DOI

10.1109/ISP.2003.1275848

Filename

1275848