State mixture modelling applied to speech recognition

Dat Tran; Michael Wagner; Tongtao Zheng

doi:10.1016/S0167-8655(99)00113-0

State mixture modelling applied to speech recognition

Dat Tran, Michael Wagner, Tongtao Zheng

Information Systems

Research output: Contribution to journal › Article › peer-review

Abstract

In state mixture modelling (SMM), the temporal structure of the observation sequences is represented by the state joint probability distribution where mixtures of states are considered. This technique is considered in an iterative scheme via maximum likelihood estimation. A fuzzy estimation approach is also introduced to cooperate with the SMM model. This new approach not only saves calculations from 2N^TT (HMM direct calculation) and N²T (Forward-backward algorithm) to just only 2NT calculations, but also achieves a better recognition result.

Original language	English
Pages (from-to)	1449-1456
Number of pages	8
Journal	Pattern Recognition Letters
Volume	20
Issue number	11-13
DOIs	https://doi.org/10.1016/S0167-8655(99)00113-0
Publication status	Published - Nov 1999
Event	Proceedings of the 1999 Pattern Recognition in Practice (PRP VI) - Vlieland, Neth Duration: 2 Jun 1999 → 4 Jun 1999

Access to Document

10.1016/S0167-8655(99)00113-0

Cite this

@article{da8ae8d27478496ab4442369dfa1abf1,

title = "State mixture modelling applied to speech recognition",

abstract = "In state mixture modelling (SMM), the temporal structure of the observation sequences is represented by the state joint probability distribution where mixtures of states are considered. This technique is considered in an iterative scheme via maximum likelihood estimation. A fuzzy estimation approach is also introduced to cooperate with the SMM model. This new approach not only saves calculations from 2NTT (HMM direct calculation) and N2T (Forward-backward algorithm) to just only 2NT calculations, but also achieves a better recognition result.",

author = "Dat Tran and Michael Wagner and Tongtao Zheng",

year = "1999",

month = nov,

doi = "10.1016/S0167-8655(99)00113-0",

language = "English",

volume = "20",

pages = "1449--1456",

journal = "Pattern Recognition Letters",

issn = "0167-8655",

publisher = "Elsevier",

number = "11-13",

note = "Proceedings of the 1999 Pattern Recognition in Practice (PRP VI) ; Conference date: 02-06-1999 Through 04-06-1999",

}

TY - JOUR

T1 - State mixture modelling applied to speech recognition

AU - Tran, Dat

AU - Wagner, Michael

AU - Zheng, Tongtao

PY - 1999/11

Y1 - 1999/11

N2 - In state mixture modelling (SMM), the temporal structure of the observation sequences is represented by the state joint probability distribution where mixtures of states are considered. This technique is considered in an iterative scheme via maximum likelihood estimation. A fuzzy estimation approach is also introduced to cooperate with the SMM model. This new approach not only saves calculations from 2NTT (HMM direct calculation) and N2T (Forward-backward algorithm) to just only 2NT calculations, but also achieves a better recognition result.

AB - In state mixture modelling (SMM), the temporal structure of the observation sequences is represented by the state joint probability distribution where mixtures of states are considered. This technique is considered in an iterative scheme via maximum likelihood estimation. A fuzzy estimation approach is also introduced to cooperate with the SMM model. This new approach not only saves calculations from 2NTT (HMM direct calculation) and N2T (Forward-backward algorithm) to just only 2NT calculations, but also achieves a better recognition result.

UR - http://www.scopus.com/inward/record.url?scp=0033220807&partnerID=8YFLogxK

U2 - 10.1016/S0167-8655(99)00113-0

DO - 10.1016/S0167-8655(99)00113-0

M3 - Article

AN - SCOPUS:0033220807

SN - 0167-8655

VL - 20

SP - 1449

EP - 1456

JO - Pattern Recognition Letters

JF - Pattern Recognition Letters

IS - 11-13

T2 - Proceedings of the 1999 Pattern Recognition in Practice (PRP VI)

Y2 - 2 June 1999 through 4 June 1999

ER -

State mixture modelling applied to speech recognition

Abstract

Access to Document

Other files and links

Cite this