Informatics Report Series



Related Pages

Report (by Number) Index
Report (by Date) Index
Author Index
Institute Index

Title:Speech recognition using linear dynamic models
Authors: Jolyon Frankel ; Simon King
Date:Jan 2007
Publication Title:IEEE Transactions on Speech and Audio Processing
Publication Type:Journal Article Publication Status:Published
Volume No:15 (1) Page Nos:246-256
DOI:10.1109/TASL.2006.876766 ISBN/ISSN:1558-7916
The majority of automatic speech recognition (ASR) systems rely on hidden Markov models, in which Gaussian mixtures model the output distributions associated with sub-phone states. This approach, whilst successful, models consecutive feature vectors (augmented to include derivative information) as statistically independent. Furthermore, spatial correlations present in speech parameters are frequently ignored through the use of diagonal covariance matrices. This paper continues the work of Digalakis and others who proposed instead a first-order linear state-space model which has the capacity to model underlying dynamics, and furthermore give a model of spatial correlations. This paper examines the assumptions made in applying such a model and shows that the addition of a hidden dynamic state leads to increases in accuracy over otherwise equivalent static models. We also propose a time-asynchronous decoding strategy suited to recognition with segment models. We describe implementation of decoding for linear dynamic models and present TIMIT phone recognition results.
Links To Paper
1st Link
Bibtex format
author = { Jolyon Frankel and Simon King },
title = {Speech recognition using linear dynamic models},
journal = {IEEE Transactions on Speech and Audio Processing},
publisher = {IEEE},
year = 2007,
month = {Jan},
volume = {15 (1)},
pages = {246-256},
doi = {10.1109/TASL.2006.876766},
url = {},

Home : Publications : Report 

Please mail <> with any changes or corrections.
Unless explicitly stated otherwise, all material is copyright The University of Edinburgh