Dynamic Cepstral Representations Based on Order-Dependent Windowing Methods

Hong Kook KIM; Seung Ho CHOI; Hwang Soo LEE

Dynamic Cepstral Representations Based on Order-Dependent Windowing Methods

Hong Kook KIM, Seung Ho CHOI, Hwang Soo LEE

Full Text Views

0

Share
Cite this

Summary :

In this paper, we propose dynamic cepstral representations to effectively capture the temporal information of cepstral coefficients. The number of speech frames for the regression analysis to extract a dynamic cepstral coefficient is inversely proportional to the cepstral order since the cepstral coefficients of higher orders are more fluctuating than those of lower orders. By exploiting the relationship between the window length for extracting a dynamic cepstral coefficient and the statistical variance of the cepstral coefficient, we propose three kinds of windowing methods in this work: an utterance-specific variance-ratio windowing method, a statistical variance-ratio windowing method, and an inverse-lifter windowing method. Intra-speaker, inter-speaker, and speaker-independent recognition tests on 100 phonetically balanced words are carried out to evaluate the performance of the proposed order-dependent windowing methods.

Publication: IEICE TRANSACTIONS on Information Vol.E81-D No.5 pp.434-440

Publication Date: 1998/05/25

Publicized

Online ISSN

DOI

Type of Manuscript

Category: Speech Processing and Acoustics

Cite this

Copy

Hong Kook KIM, Seung Ho CHOI, Hwang Soo LEE, "Dynamic Cepstral Representations Based on Order-Dependent Windowing Methods" in IEICE TRANSACTIONS on Information, vol. E81-D, no. 5, pp. 434-440, May 1998, doi: .
Abstract: In this paper, we propose dynamic cepstral representations to effectively capture the temporal information of cepstral coefficients. The number of speech frames for the regression analysis to extract a dynamic cepstral coefficient is inversely proportional to the cepstral order since the cepstral coefficients of higher orders are more fluctuating than those of lower orders. By exploiting the relationship between the window length for extracting a dynamic cepstral coefficient and the statistical variance of the cepstral coefficient, we propose three kinds of windowing methods in this work: an utterance-specific variance-ratio windowing method, a statistical variance-ratio windowing method, and an inverse-lifter windowing method. Intra-speaker, inter-speaker, and speaker-independent recognition tests on 100 phonetically balanced words are carried out to evaluate the performance of the proposed order-dependent windowing methods.
URL: https://globals.ieice.org/en_transactions/information/10.1587/e81-d_5_434/_p

Copy

@ARTICLE{e81-d_5_434,
author={Hong Kook KIM, Seung Ho CHOI, Hwang Soo LEE, },
journal={IEICE TRANSACTIONS on Information},
title={Dynamic Cepstral Representations Based on Order-Dependent Windowing Methods},
year={1998},
volume={E81-D},
number={5},
pages={434-440},
abstract={In this paper, we propose dynamic cepstral representations to effectively capture the temporal information of cepstral coefficients. The number of speech frames for the regression analysis to extract a dynamic cepstral coefficient is inversely proportional to the cepstral order since the cepstral coefficients of higher orders are more fluctuating than those of lower orders. By exploiting the relationship between the window length for extracting a dynamic cepstral coefficient and the statistical variance of the cepstral coefficient, we propose three kinds of windowing methods in this work: an utterance-specific variance-ratio windowing method, a statistical variance-ratio windowing method, and an inverse-lifter windowing method. Intra-speaker, inter-speaker, and speaker-independent recognition tests on 100 phonetically balanced words are carried out to evaluate the performance of the proposed order-dependent windowing methods.},
keywords={},
doi={},
ISSN={},
month={May},}

Copy

TY - JOUR
TI - Dynamic Cepstral Representations Based on Order-Dependent Windowing Methods
T2 - IEICE TRANSACTIONS on Information
SP - 434
EP - 440
AU - Hong Kook KIM
AU - Seung Ho CHOI
AU - Hwang Soo LEE
PY - 1998
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E81-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 1998
AB - In this paper, we propose dynamic cepstral representations to effectively capture the temporal information of cepstral coefficients. The number of speech frames for the regression analysis to extract a dynamic cepstral coefficient is inversely proportional to the cepstral order since the cepstral coefficients of higher orders are more fluctuating than those of lower orders. By exploiting the relationship between the window length for extracting a dynamic cepstral coefficient and the statistical variance of the cepstral coefficient, we propose three kinds of windowing methods in this work: an utterance-specific variance-ratio windowing method, a statistical variance-ratio windowing method, and an inverse-lifter windowing method. Intra-speaker, inter-speaker, and speaker-independent recognition tests on 100 phonetically balanced words are carried out to evaluate the performance of the proposed order-dependent windowing methods.
ER -