Catalogo Articoli (Spogli Riviste)

OPAC HELP

Titolo:
Noise robust speech parameterization using multiresolution feature extraction
Autore:
Hariharan, R; Kiss, I; Viikki, O;
Indirizzi:
Nokia Res Ctr, Speech & Audio Syst Lab, FIN-33721 Tampere, Finland Nokia Res Ctr Tampere Finland FIN-33721 Lab, FIN-33721 Tampere, Finland
Titolo Testata:
IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING
fascicolo: 8, volume: 9, anno: 2001,
pagine: 856 - 865
SICI:
1063-6676(200111)9:8<856:NRSPUM>2.0.ZU;2-4
Fonte:
ISI
Lingua:
ENG
Soggetto:
HIDDEN MARKOV-MODELS; RECOGNITION;
Keywords:
cesptral normalization; feature extraction; front-end; multiband speech recognition; multiresolution front-end; noise robust feature extraction; noise robustness; speech recognition; subband speech recognition;
Tipo documento:
Article
Natura:
Periodico
Settore Disciplinare:
Engineering, Computing & Technology
Citazioni:
34
Recensione:
Indirizzi per estratti:
Indirizzo: Hariharan, R Nokia Res Ctr, Speech & Audio Syst Lab, FIN-33721 Tampere, Finland Nokia Res Ctr Tampere Finland FIN-33721 21 Tampere, Finland
Citazione:
R. Hariharan et al., "Noise robust speech parameterization using multiresolution feature extraction", IEEE SPEECH, 9(8), 2001, pp. 856-865

Abstract

In this paper, we present a multiresolution-based feature extraction technique for speech recognition in adverse conditions. The proposed front-end algorithm uses mel cepstrum-based feature computation of subbands in order not to spread noise distortions over the entire feature space. Conventional full-band features are also augmented to the final feature vector which is fed to the recognition unit. Other novel features of the proposed front-endalgorithm include emphasis of long-term spectral information combined withcepstral domain feature vector normalization and the use of the PCA transform, instead of DCT, to provide the final cepstral parameters. The proposedalgorithm was experimentally evaluated in a connected digit recognition task under various noise conditions. The results obtained show that the new feature extraction algorithm improves word recognition accuracy by 41% when compared to the performance of mel cepstrum front-end. A substantial increase in recognition accuracy was observed in all tested noise environments atall different SNRs. The good performance of multiresolution front-end is not only due to the higher feature vector dimension, but the proposed algorithm clearly outperformed the mel cepstral front-end when the same number ofHMM parameters were used in both systems. We also propose methods to reduce the computational complexity of the multiresolution front-end-based speech recognition system. Experimental results indicate the viability of the proposed techniques.

ASDD Area Sistemi Dipartimentali e Documentali, Università di Bologna, Catalogo delle riviste ed altri periodici
Documento generato il 03/04/20 alle ore 04:27:59