TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS

Phil Weber; Stephen Houghton; Colin Champion; Martin Russell; Peter Jancovic

TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS

Phil Weber, Stephen Houghton, Colin Champion, Martin Russell, Peter Jancovic

Electronic, Electrical and Systems Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

6 Citations (Scopus)

Abstract

Many current speech models used in recognition involve thousands of parameters, whereas the mechanisms of speech production are conceptually very simple. We present and evaluate a new continuous state probabilistic model (CS-HMM) for recovering dwell-transition and phoneme sequences from dynamic speech production features. We show that with very few parameters, these features can be tracked, and phoneme sequences recovered, with promising accuracy.

Original language	English
Title of host publication	2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP)
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	3042-3046
ISBN (Electronic)	978-1-4799-2893-4
Publication status	Published - 2014
Event	2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014 - Florence, Italy, Florence, Italy Duration: 4 May 2014 → 9 May 2014

Conference

Conference	2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014
Country/Territory	Italy
City	Florence
Period	4/05/14 → 9/05/14

Speech Recognition by Synthesis (SRbS)
Russell, M. & Jancovic, P.
1/10/12 → 31/10/16
Project: Other Government Departments

Cite this

@inproceedings{962ac96a7e4344a3a5a31b1e517d64ae,

title = "TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS",

abstract = "Many current speech models used in recognition involve thousands of parameters, whereas the mechanisms of speech production are conceptually very simple. We present and evaluate a new continuous state probabilistic model (CS-HMM) for recovering dwell-transition and phoneme sequences from dynamic speech production features. We show that with very few parameters, these features can be tracked, and phoneme sequences recovered, with promising accuracy.",

author = "Phil Weber and Stephen Houghton and Colin Champion and Martin Russell and Peter Jancovic",

year = "2014",

language = "English",

pages = "3042--3046",

booktitle = "2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

note = "2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014 ; Conference date: 04-05-2014 Through 09-05-2014",

}

Weber, P, Houghton, S, Champion, C, Russell, M & Jancovic, P 2014, TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS. in 2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP). Institute of Electrical and Electronics Engineers (IEEE), pp. 3042-3046, 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014, Florence, Italy, 4/05/14.

TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS. / Weber, Phil; Houghton, Stephen; Champion, Colin et al.
2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP). Institute of Electrical and Electronics Engineers (IEEE), 2014. p. 3042-3046.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS

AU - Weber, Phil

AU - Houghton, Stephen

AU - Champion, Colin

AU - Russell, Martin

AU - Jancovic, Peter

PY - 2014

Y1 - 2014

N2 - Many current speech models used in recognition involve thousands of parameters, whereas the mechanisms of speech production are conceptually very simple. We present and evaluate a new continuous state probabilistic model (CS-HMM) for recovering dwell-transition and phoneme sequences from dynamic speech production features. We show that with very few parameters, these features can be tracked, and phoneme sequences recovered, with promising accuracy.

AB - Many current speech models used in recognition involve thousands of parameters, whereas the mechanisms of speech production are conceptually very simple. We present and evaluate a new continuous state probabilistic model (CS-HMM) for recovering dwell-transition and phoneme sequences from dynamic speech production features. We show that with very few parameters, these features can be tracked, and phoneme sequences recovered, with promising accuracy.

M3 - Conference contribution

SP - 3042

EP - 3046

BT - 2014 IEEE International Conference on Acoustic, Speech and Signal Processing (ICASSP)

PB - Institute of Electrical and Electronics Engineers (IEEE)

T2 - 2014 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2014

Y2 - 4 May 2014 through 9 May 2014

ER -

TRAJECTORY ANALYSIS OF SPEECH USING CONTINUOUS STATE HIDDEN MARKOV MODELS

Abstract

Conference

Fingerprint

Projects

Speech Recognition by Synthesis (SRbS)

Cite this