On the use of the correlation between acoustic descriptors for the normal/pathological voices discrimination

Authors:
Thomas Dubuisson;Thierry Dutoit;Bernard Gosselin;Marc Remacle
Affiliations:
TCTS Lab, Faculté Polytechnique de Mons, Mons, Belgium;TCTS Lab, Faculté Polytechnique de Mons, Mons, Belgium;TCTS Lab, Faculté Polytechnique de Mons, Mons, Belgium;ORL, ORLO Lab, Université Catholique de Louvain, Yvoir, Belgium
Venue:
EURASIP Journal on Advances in Signal Processing - Special issue on analysis and signal processing of oesophageal and pathological voices
Year:
2009

Citing 8
Cited 1

Introduction to statistical pattern recognition (2nd ed.)

Introduction to statistical pattern recognition (2nd ed.)
Automatic detection of pathologies in the voice by HOS based parameters

EURASIP Journal on Applied Signal Processing - Nonlinear signal and image processing - part I
Content-Based Classification, Search, and Retrieval of Audio

IEEE MultiMedia
Construction and Evaluation of a Robust Multifeature Speech/Music Discriminator

ICASSP '97 Proceedings of the 1997 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP '97)-Volume 2 - Volume 2
Pattern Recognition and Machine Learning (Information Science and Statistics)

Pattern Recognition and Machine Learning (Information Science and Statistics)
Study of harmonics-to-noise ratio and critical-band energy spectrum of speech as acoustic indicators of laryngeal and voice pathology

EURASIP Journal on Applied Signal Processing
Multiple scale music segmentation using rhythm, timbre, and harmony

EURASIP Journal on Applied Signal Processing
Springer Handbook of Speech Processing

Springer Handbook of Speech Processing

An improved method for voice pathology detection by means of a HMM-based feature space transformation

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents an analysis system aiming at discriminating between normal and pathological voices. Compared to literature of voice pathology assessment, it is characterised by two aspects. First the system is based on features inspired from voice pathology assessment and music information retrieval. Second the distinction between normal and pathological voices is simply based on the correlation between acoustic features, while more complex classifiers are common in literature. Based on the normal and pathological samples included the MEEI database, it has been found that using two features (spectral decrease and first spectral tristimulus in the Bark scale) and their correlation leads to correct classification rates of 94.7% for pathological voices and 89.5% for normal ones. The system also outputs a normal/pathological factor aiming at giving an indication to the clinician about the location of a subject according to the database.