The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
Este artigo propõe um novo tipo de modelo oculto de Markov (HMM) baseado na distribuição de probabilidade multi-espaço e deriva um algoritmo de estimativa de parâmetros para o HMM estendido. HMMs são modelos estatísticos amplamente utilizados para caracterizar sequências de espectros de fala e têm sido aplicados com sucesso em sistemas de reconhecimento de fala. Os HMMs são categorizados em HMMs discretos e HMMs contínuos, que podem modelar sequências de símbolos discretos e vetores contínuos, respectivamente. No entanto, não podemos aplicar os HMMs convencionais discretos e contínuos a sequências de observação que consistem em valores contínuos e símbolos discretos: a modelagem do padrão F0 da fala é uma boa ilustração. O HMM proposto inclui HMM discreto e HMM contínuo como casos especiais e, além disso, pode modelar sequências que consistem em vetores de observação com dimensionalidade variável e símbolos discretos.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copiar
Keiichi TOKUDA, Takashi MASUKO, Noboru MIYAZAKI, Takao KOBAYASHI, "Multi-Space Probability Distribution HMM" in IEICE TRANSACTIONS on Information,
vol. E85-D, no. 3, pp. 455-464, March 2002, doi: .
Abstract: This paper proposes a new kind of hidden Markov model (HMM) based on multi-space probability distribution, and derives a parameter estimation algorithm for the extended HMM. HMMs are widely used statistical models for characterizing sequences of speech spectra, and have been successfully applied to speech recognition systems. HMMs are categorized into discrete HMMs and continuous HMMs, which can model sequences of discrete symbols and continuous vectors, respectively. However, we cannot apply both the conventional discrete and continuous HMMs to observation sequences which consist of continuous values and discrete symbols: F0 pattern modeling of speech is a good illustration. The proposed HMM includes discrete HMM and continuous HMM as special cases, and furthermore, can model sequences which consist of observation vectors with variable dimensionality and discrete symbols.
URL: https://global.ieice.org/en_transactions/information/10.1587/e85-d_3_455/_p
Copiar
@ARTICLE{e85-d_3_455,
author={Keiichi TOKUDA, Takashi MASUKO, Noboru MIYAZAKI, Takao KOBAYASHI, },
journal={IEICE TRANSACTIONS on Information},
title={Multi-Space Probability Distribution HMM},
year={2002},
volume={E85-D},
number={3},
pages={455-464},
abstract={This paper proposes a new kind of hidden Markov model (HMM) based on multi-space probability distribution, and derives a parameter estimation algorithm for the extended HMM. HMMs are widely used statistical models for characterizing sequences of speech spectra, and have been successfully applied to speech recognition systems. HMMs are categorized into discrete HMMs and continuous HMMs, which can model sequences of discrete symbols and continuous vectors, respectively. However, we cannot apply both the conventional discrete and continuous HMMs to observation sequences which consist of continuous values and discrete symbols: F0 pattern modeling of speech is a good illustration. The proposed HMM includes discrete HMM and continuous HMM as special cases, and furthermore, can model sequences which consist of observation vectors with variable dimensionality and discrete symbols.},
keywords={},
doi={},
ISSN={},
month={March},}
Copiar
TY - JOUR
TI - Multi-Space Probability Distribution HMM
T2 - IEICE TRANSACTIONS on Information
SP - 455
EP - 464
AU - Keiichi TOKUDA
AU - Takashi MASUKO
AU - Noboru MIYAZAKI
AU - Takao KOBAYASHI
PY - 2002
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E85-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2002
AB - This paper proposes a new kind of hidden Markov model (HMM) based on multi-space probability distribution, and derives a parameter estimation algorithm for the extended HMM. HMMs are widely used statistical models for characterizing sequences of speech spectra, and have been successfully applied to speech recognition systems. HMMs are categorized into discrete HMMs and continuous HMMs, which can model sequences of discrete symbols and continuous vectors, respectively. However, we cannot apply both the conventional discrete and continuous HMMs to observation sequences which consist of continuous values and discrete symbols: F0 pattern modeling of speech is a good illustration. The proposed HMM includes discrete HMM and continuous HMM as special cases, and furthermore, can model sequences which consist of observation vectors with variable dimensionality and discrete symbols.
ER -