Optimization of perceptually-based ASR front-end [automatic speech recognition]

Author

Hermansky, Hynek ; Junqua, Jean Claude

Author_Institution

Speech Technol. Lab., Santa Barbara, CA, USA

fYear

1988

fDate

11-14 Apr 1988

Firstpage

219

Abstract

Several recently proposed automatic speech recognition (ASR) front-ends are experimentally compared for speaker-dependent and cross-speaker ASR. The perceptually based linear predictive front-end yields the highest accuracies. By modifying its sensitivity to spectral peaks and to spectral tilt and by utilizing the speech dynamics the authors further improve, by about 10%, its error rate in speaker-independent ASR

Keywords

speech recognition; automatic speech recognition; cross-speaker ASR; error rate; optimisation; perceptually based linear predictive front-end; sensitivity; speaker-dependent ASR; speaker-independent ASR; spectral peaks; spectral tilt; speech dynamics; Automatic speech recognition; Bandwidth; Cepstral analysis; Databases; Filters; Frequency; Laboratories; Predictive models; Psychology; Speech analysis;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1988. ICASSP-88., 1988 International Conference on

Conference_Location

New York, NY

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1988.196553

Filename

196553

Link To Document

https://search.isc.ac/dl/search/defaultta.aspx?DTC=49&DC=2997442