Experiments with fast Fourier transform, linear predictive and cepstral coefficients in dysarthric speech recognition algorithms using hidden Markov Model.
P. Polur, and G. Miller. IEEE Trans Neural Syst Rehabil Eng, 13 (4):
558--561(December 2005)
Abstract
In this study, a hidden Markov Model was constructed and conditions were investigated that would provide improved performance for a dysarthric speech (isolated word) recognition system. The speaker dependant system was intended to act as an assistive/control tool. A small size vocabulary spoken by three cerebral palsy subjects was chosen. Fast Fourier transform, linear predictive, and Mel frequency cepstral coefficients extracted from data provided training input to several whole-word hidden Markov model configurations. The effect of model structure, number of states, and frame rates were also investigated. It was noted that a 10-state ergodic model using 15 msec frames was better than other configurations. Furthermore, it was found that a Mel cepstrum based model outperformed a fast Fourier transform and linear prediction based model. The system offers effective and robust application as a rehabilitation and/or control tool to assist dysarthric motor impaired individuals.
%0 Journal Article
%1 Polur2005
%A Polur, Prasad D
%A Miller, Gerald E
%D 2005
%J IEEE Trans Neural Syst Rehabil Eng
%K Algorithms; Art; Cerebral Palsy; Dysarthria; Fourier Analysis; Humans; Linear Models; Markov Chains; Models, Biological; Pattern Recognition, Automated; Signal Processing, Computer-Assisted; Speech Recognition Software; ificial Intelligence
%N 4
%P 558--561
%T Experiments with fast Fourier transform, linear predictive and cepstral coefficients in dysarthric speech recognition algorithms using hidden Markov Model.
%V 13
%X In this study, a hidden Markov Model was constructed and conditions were investigated that would provide improved performance for a dysarthric speech (isolated word) recognition system. The speaker dependant system was intended to act as an assistive/control tool. A small size vocabulary spoken by three cerebral palsy subjects was chosen. Fast Fourier transform, linear predictive, and Mel frequency cepstral coefficients extracted from data provided training input to several whole-word hidden Markov model configurations. The effect of model structure, number of states, and frame rates were also investigated. It was noted that a 10-state ergodic model using 15 msec frames was better than other configurations. Furthermore, it was found that a Mel cepstrum based model outperformed a fast Fourier transform and linear prediction based model. The system offers effective and robust application as a rehabilitation and/or control tool to assist dysarthric motor impaired individuals.
@article{Polur2005,
abstract = {In this study, a hidden Markov Model was constructed and conditions were investigated that would provide improved performance for a dysarthric speech (isolated word) recognition system. The speaker dependant system was intended to act as an assistive/control tool. A small size vocabulary spoken by three cerebral palsy subjects was chosen. Fast Fourier transform, linear predictive, and Mel frequency cepstral coefficients extracted from data provided training input to several whole-word hidden Markov model configurations. The effect of model structure, number of states, and frame rates were also investigated. It was noted that a 10-state ergodic model using 15 msec frames was better than other configurations. Furthermore, it was found that a Mel cepstrum based model outperformed a fast Fourier transform and linear prediction based model. The system offers effective and robust application as a rehabilitation and/or control tool to assist dysarthric motor impaired individuals.},
added-at = {2014-07-19T21:03:42.000+0200},
author = {Polur, Prasad D and Miller, Gerald E},
biburl = {https://www.bibsonomy.org/bibtex/2b64d8c59401eed3a66421bada9246cd9/ar0berts},
groups = {public},
interhash = {93c13c4b8b25cfa12baf1a2bdc07687b},
intrahash = {b64d8c59401eed3a66421bada9246cd9},
journal = {IEEE Trans Neural Syst Rehabil Eng},
keywords = {Algorithms; Art; Cerebral Palsy; Dysarthria; Fourier Analysis; Humans; Linear Models; Markov Chains; Models, Biological; Pattern Recognition, Automated; Signal Processing, Computer-Assisted; Speech Recognition Software; ificial Intelligence},
month = Dec,
number = 4,
pages = {558--561},
pmid = {16425838},
timestamp = {2014-07-19T21:03:42.000+0200},
title = {Experiments with fast Fourier transform, linear predictive and cepstral coefficients in dysarthric speech recognition algorithms using hidden Markov Model.},
username = {ar0berts},
volume = 13,
year = 2005
}