Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process

Toru TAKAHASHI; Keiichi TOKUDA; Takao KOBAYASHI; Tadashi KITAMURA

Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process

Toru TAKAHASHI, Keiichi TOKUDA, Takao KOBAYASHI, Tadashi KITAMURA

Full Text Views

0

Cite this

Summary :

This paper defines a new kind of a mixture density model for modeling a quasi-stationary Gaussian process based on mel-cepstral representation. The conventional AR mixture density model can be applied to modeling a quasi-stationary Gaussian AR process. However, it cannot model spectral zeros. In contrast, the proposed model is based on a frequency-warped exponential (EX) model. Accordingly, it can represent spectral poles and zeros with equal weights, and, furthermore, the model spectrum has a high resolution at low frequencies. The parameter estimation algorithm for the proposed model was also derived based on an EM algorithm. Experimental results show that the proposed model has better performance than the AR mixture density model for modeling a frequency-warped EX process.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E86-A No.8 pp.1971-1978

Publication Date: 2003/08/01

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section PAPER (Special Section on Digital Signal Processing)

Category

Cite this

Copy

Toru TAKAHASHI, Keiichi TOKUDA, Takao KOBAYASHI, Tadashi KITAMURA, "Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process" in IEICE TRANSACTIONS on Fundamentals, vol. E86-A, no. 8, pp. 1971-1978, August 2003, doi: .
Abstract: This paper defines a new kind of a mixture density model for modeling a quasi-stationary Gaussian process based on mel-cepstral representation. The conventional AR mixture density model can be applied to modeling a quasi-stationary Gaussian AR process. However, it cannot model spectral zeros. In contrast, the proposed model is based on a frequency-warped exponential (EX) model. Accordingly, it can represent spectral poles and zeros with equal weights, and, furthermore, the model spectrum has a high resolution at low frequencies. The parameter estimation algorithm for the proposed model was also derived based on an EM algorithm. Experimental results show that the proposed model has better performance than the AR mixture density model for modeling a frequency-warped EX process.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/e86-a_8_1971/_p

Copy

@ARTICLE{e86-a_8_1971,
author={Toru TAKAHASHI, Keiichi TOKUDA, Takao KOBAYASHI, Tadashi KITAMURA, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process},
year={2003},
volume={E86-A},
number={8},
pages={1971-1978},
abstract={This paper defines a new kind of a mixture density model for modeling a quasi-stationary Gaussian process based on mel-cepstral representation. The conventional AR mixture density model can be applied to modeling a quasi-stationary Gaussian AR process. However, it cannot model spectral zeros. In contrast, the proposed model is based on a frequency-warped exponential (EX) model. Accordingly, it can represent spectral poles and zeros with equal weights, and, furthermore, the model spectrum has a high resolution at low frequencies. The parameter estimation algorithm for the proposed model was also derived based on an EM algorithm. Experimental results show that the proposed model has better performance than the AR mixture density model for modeling a frequency-warped EX process.},
keywords={},
doi={},
ISSN={},
month={August},}

Copy

TY - JOUR
TI - Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 1971
EP - 1978
AU - Toru TAKAHASHI
AU - Keiichi TOKUDA
AU - Takao KOBAYASHI
AU - Tadashi KITAMURA
PY - 2003
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E86-A
IS - 8
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - August 2003
AB - This paper defines a new kind of a mixture density model for modeling a quasi-stationary Gaussian process based on mel-cepstral representation. The conventional AR mixture density model can be applied to modeling a quasi-stationary Gaussian AR process. However, it cannot model spectral zeros. In contrast, the proposed model is based on a frequency-warped exponential (EX) model. Accordingly, it can represent spectral poles and zeros with equal weights, and, furthermore, the model spectrum has a high resolution at low frequencies. The parameter estimation algorithm for the proposed model was also derived based on an EM algorithm. Experimental results show that the proposed model has better performance than the AR mixture density model for modeling a frequency-warped EX process.
ER -