Use of periodicity and jitter as speech recognition features

Author

Thomson, David L. ; Chengalvarayan, Rathinavelu

Author_Institution

Lucent Technol., AT&T Bell Labs., Naperville, IL, USA

Volume

1

fYear

1998

fDate

12-15 May 1998

Firstpage

21

Abstract

We investigate a class of features related to voicing parameters that indicate whether the vocal chords are vibrating. Features describing voicing characteristics of speech signals are integrated with an existing 38-dimensional feature vector consisting of first and second order time derivatives of the frame energy and of the cepstral coefficients with their first and second derivatives. HMM-based connected digit recognition experiments comparing the traditional and extended feature sets show that voicing features and spectral information are complementary and that improved speech recognition performance is obtained by combining the two sources of information

Keywords

cepstral analysis; feature extraction; hidden Markov models; jitter; maximum likelihood estimation; parameter estimation; signal representation; speech processing; speech recognition; HMM-based connected digit recognition; cepstral coefficients; experiments; extended feature sets; first order time derivative; frame energy; jitter; maximum likelihood method; minimum string error; periodicity; second order time derivative; signal representation; speech recognition features; speech recognition performance; speech signals; training method; vocal chords; voicing parameters; Autocorrelation; Cepstral analysis; Error analysis; Information resources; Jitter; Robustness; Signal representations; Speech coding; Speech processing; Speech recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on

Conference_Location

Seattle, WA

ISSN

1520-6149

Print_ISBN

0-7803-4428-6

Type

conf

DOI

10.1109/ICASSP.1998.674357

Filename

674357