An investigation into the correlation and prediction of acoustic speech features from MFCC vectors

Jonathan Darch; Ben P. Milner; Ibrahim Almajai; Saeed V. Vaseghi

doi:10.1109/ICASSP.2007.366950

An investigation into the correlation and prediction of acoustic speech features from MFCC vectors

Jonathan Darch, Ben P. Milner, Ibrahim Almajai, Saeed V. Vaseghi

Research output: Contribution to conference › Paper

4 Citations (Scopus)

Abstract

This work develops a statistical framework to predict acoustic features (fundamental frequency, formant frequencies and voicing) from MFCC vectors. An analysis of correlation between acoustic features and MFCCs is made both globally across all speech and within phoneme classes, and also from speaker-independent and speaker-dependent speech. This leads to the development of both a global prediction method, using a Gaussian mixture model (GMM) to model the joint density of acoustic features and MFCCs, and a phoneme-specific prediction method using a combined hidden Markov model (HMM)-GMM. Prediction accuracy measurements show the phoneme-dependent HMM-GMM system to be more accurate which agrees with the correlation analysis. Results also show prediction to be more accurate from speaker-dependent speech which also agrees with the correlation analysis

Original language	English
Pages	IV-465-IV-468
Number of pages	4
DOIs	https://doi.org/10.1109/ICASSP.2007.366950
Publication status	Published - 2007
Event	IEEE International Conference on Acoustics, Speech and Signal Processing - Honolulu, United States Duration: 15 Apr 2007 → 20 Apr 2007

Conference

Conference	IEEE International Conference on Acoustics, Speech and Signal Processing
Country/Territory	United States
City	Honolulu
Period	15/04/07 → 20/04/07

Access to Document

10.1109/ICASSP.2007.366950

Cite this

@conference{eca0b5fba0b940429dd047fce04311b6,

title = "An investigation into the correlation and prediction of acoustic speech features from MFCC vectors",

abstract = "This work develops a statistical framework to predict acoustic features (fundamental frequency, formant frequencies and voicing) from MFCC vectors. An analysis of correlation between acoustic features and MFCCs is made both globally across all speech and within phoneme classes, and also from speaker-independent and speaker-dependent speech. This leads to the development of both a global prediction method, using a Gaussian mixture model (GMM) to model the joint density of acoustic features and MFCCs, and a phoneme-specific prediction method using a combined hidden Markov model (HMM)-GMM. Prediction accuracy measurements show the phoneme-dependent HMM-GMM system to be more accurate which agrees with the correlation analysis. Results also show prediction to be more accurate from speaker-dependent speech which also agrees with the correlation analysis",

author = "Jonathan Darch and Milner, {Ben P.} and Ibrahim Almajai and Vaseghi, {Saeed V.}",

year = "2007",

doi = "10.1109/ICASSP.2007.366950",

language = "English",

pages = "IV--465--IV--468",

note = "IEEE International Conference on Acoustics, Speech and Signal Processing ; Conference date: 15-04-2007 Through 20-04-2007",

}

TY - CONF

T1 - An investigation into the correlation and prediction of acoustic speech features from MFCC vectors

AU - Darch, Jonathan

AU - Milner, Ben P.

AU - Almajai, Ibrahim

AU - Vaseghi, Saeed V.

PY - 2007

Y1 - 2007

N2 - This work develops a statistical framework to predict acoustic features (fundamental frequency, formant frequencies and voicing) from MFCC vectors. An analysis of correlation between acoustic features and MFCCs is made both globally across all speech and within phoneme classes, and also from speaker-independent and speaker-dependent speech. This leads to the development of both a global prediction method, using a Gaussian mixture model (GMM) to model the joint density of acoustic features and MFCCs, and a phoneme-specific prediction method using a combined hidden Markov model (HMM)-GMM. Prediction accuracy measurements show the phoneme-dependent HMM-GMM system to be more accurate which agrees with the correlation analysis. Results also show prediction to be more accurate from speaker-dependent speech which also agrees with the correlation analysis

AB - This work develops a statistical framework to predict acoustic features (fundamental frequency, formant frequencies and voicing) from MFCC vectors. An analysis of correlation between acoustic features and MFCCs is made both globally across all speech and within phoneme classes, and also from speaker-independent and speaker-dependent speech. This leads to the development of both a global prediction method, using a Gaussian mixture model (GMM) to model the joint density of acoustic features and MFCCs, and a phoneme-specific prediction method using a combined hidden Markov model (HMM)-GMM. Prediction accuracy measurements show the phoneme-dependent HMM-GMM system to be more accurate which agrees with the correlation analysis. Results also show prediction to be more accurate from speaker-dependent speech which also agrees with the correlation analysis

U2 - 10.1109/ICASSP.2007.366950

DO - 10.1109/ICASSP.2007.366950

M3 - Paper

SP - IV-465-IV-468

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

Y2 - 15 April 2007 through 20 April 2007

ER -