Acoustic analysis of the unvoiced stop consonants for detecting hypernasal speech

Abstract

Speakers having evidence of a defective velopharyngeal mechanism produce speech with inappropriate nasal resonance (hypernasal speech). Voice analysis methods for the detection of hypernasality commonly use vowels and nasalized vowels. However, to obtain a more general assessment of this abnormality it is necessary to analyze stops and fricatives. This study describes a method for hipernasality detection analyzing the unvoiced Spanish stop consonants /k/ and /p/, as well. The importance of phonemeby- phoneme analysis is shown, in contrast with whole word parametrization which may include irrelevant segments from the classification point of view. Parameters that correlate the imprints of Velopharyngeal Incompetence (VPI) over voiceless stop consonants were used in the feature estimation stage. Classification was carried out using a Support Vector Machine (SVM), obtaining a performance of 74% for a repeated cross-validation strategy evaluation.