Title :
Automatic determination of tone period and evaluation of dysphony in pathological voices
Author :
Feij¿¿o, S. ; Hern¿¿ndez, C.
Author_Institution :
University of Santiago, Department of Electronics, Faculty of Physics, Santiago de Compostela, Spain
fDate :
3/1/1986 12:00:00 AM
Abstract :
The hoarseness of 28 subjects with laryngeal complaints of widely differing severity has been automatically evaluated by using the simple Chebychev norm to compare a 3-component vector representing the subject´s voice with a normal template. A second procedure used the absolute values of the parameters without comparison with a standard. The three components of the characteristic vectors were extracted from recorded utterances of the Spanish vowel `a¿¿, for which it was necessary to determine the approximate tone period of the signals. Two methods of pitch extraction were used, one based on the autocorrelation function and the other on a SIFT algorithm.
Keywords :
computerised signal processing; medical diagnostic computing; speech; Chebychev norm; SIFT algorithm; Spanish vowel; autocorrelation function; computerised signal processing; dysphony; hoarseness; laryngeal; larynx; medical diagnostic computing; pathological voices; pitch extraction; speech; tone period;
Journal_Title :
Physical Science, Measurement and Instrumentation, Management and Education - Reviews, IEE Proceedings A
DOI :
10.1049/ip-a-1.1986.0016