Noisy speech recognition system for car cellular phone

Author

Sankar, Ravi ; Sethi, Netoo S.

Author_Institution

Dept. of Electr. Eng., Univ. of South Florida, Tampa, FL, USA

Volume

4

fYear

1999

fDate

1999

Firstpage

2218

Abstract

In this paper we present the development of robust speech recognition techniques for voice activated dialing (VAD) for a cellular phone in a car. Noise reduction techniques are implemented before modeling the speech parameters in the homomorphic domain. The FFT derived cepstral coefficients are lifted and then Mel-scale warped to generate the feature vector. A radial basis function neural network is used as final classifier and the variation in its structure is studied to evaluate its real-time performance. Results on speaker-dependent and speaker-independent recognition across -5 dB to 25 dB SNR range are presented and a performance evaluation is carried out for different front-end techniques. The system is evaluated using the NOISEX-92 database

Keywords

cellular radio; cepstral analysis; fast Fourier transforms; noise abatement; radial basis function networks; speech recognition; FFT derived cepstral coefficients; NOISEX-92 database; SNR range; VAD; car cellular phone; feature vector; front-end techniques; homomorphic domain; noise reduction techniques; noisy speech recognition system; radial basis function neural network; real-time performance; robust speech recognition techniques; speaker-dependent recognition; speaker-independent recognition; speech parameters; voice activated dialing; Cellular phones; Cepstral analysis; Neural networks; Noise reduction; Noise robustness; Radial basis function networks; Signal to noise ratio; Speech enhancement; Speech recognition; Working environment noise;

fLanguage

English

Publisher

ieee

Conference_Titel

Vehicular Technology Conference, 1999. VTC 1999 - Fall. IEEE VTS 50th

Conference_Location

Amsterdam

ISSN

1090-3038

Print_ISBN

0-7803-5435-4

Type

conf

DOI

10.1109/VETECF.1999.797332

Filename

797332