Open Access. Powered by Scholars. Published by Universities.®

Medicine and Health Sciences Commons

Open Access. Powered by Scholars. Published by Universities.®

Communication Sciences and Disorders

Western Michigan University

Masters Theses

1991

Articles 1 - 2 of 2

Full-Text Articles in Medicine and Health Sciences

Acoustic And Perceptual Correlates Of Breathy Vocal Quality, Ronald Allen Cleveland Dec 1991

Acoustic And Perceptual Correlates Of Breathy Vocal Quality, Ronald Allen Cleveland

Masters Theses

Recordings were made of seven normal female and eight normal male talkers producing sustained vowels under normal, moderately breathy, and very breathy conditions. Twenty listeners judged the recorded vowel tokens for degree of breathiness using a direct magnitude estimation procedure. A Cronbach coefficient alpha revealed strong intrasubject agreement.

Several acoustic analyses were evaluated by measuring their correlations with the mean of the listeners' breathiness ratings. Measures of cepstral peak prominence in band limited signals were most strongly correlated with perceived breathiness. The height of the autocorrelation peak in highpass filtered signals and the relative amplitude of the first harmonic were …


Time Delay Neural Networks And Speech Recognition: Context Independence Of Stops In Different Vowel Environments, Gregory Andrew Makowski Jun 1991

Time Delay Neural Networks And Speech Recognition: Context Independence Of Stops In Different Vowel Environments, Gregory Andrew Makowski

Masters Theses

A series of speech recognition experiments was conducted to investigate time-dynamic speech recognition of stop consonants invariant of vowel environment using data from six talkers. The speech preprocessing was based on previous studies investigating acoustic characteristics which correlate to the place of articulation (Blumstein and Stevens 1979). The place of articulation features were statistically abstracted using four moments and the energy level of the speech sample.

Both statistical and neural network pattern recognition methods were used. Statistical methods included linear and quadratic discriminant functions, maximum likelihood estimator (MLE) and K-nearest neighbors (KNN). The neural network approach used was Time Delay …