Vowel Classification Based on Fundamental Frequency and Formant Frequencies

Abstract

A quadratic discriminant classification technique was used to classify spectral measurements from vowels spoken by men, women, and children. The parameters used to train the discriminant classifier consisted of various combinations of fundamental frequency and the three lowest formant frequencies. Several nonlinear auditory transforms were evaluated. Unlike previous studies using a linear discriminant classifier, there was no advantage in category separability for any of the nonlinear auditory transforms over a linear frequency scale, and no advantage for spectral distances over absolute frequencies. However, it was found that parameter sets using nonlinear transforms and spectral differences reduced the differences between phonetically equivalent tokens produced by different groups of talkers.

Cross-dialectal variation in formant dynamics of American English vowels

“…For example, work by Strange and Jenkins ͑e.g., Jenkins et al, 1983Jenkins et al, , 1994Strange, 1987Strange, , 1989Strange et al, 1976 underscored the role of consonantal transitions proposing that a vowel can be reliably identified even if its "center" has been removed experimentally from the signal. On the other hand, studies by Nearey and colleagues ͑e.g., Hillenbrand and Gayvert, 1993;Hillenbrand and Nearey, 1999;Nearey and Assmann, 1986;Kewley-Port and Neel, 2006͒ demonstrated that listeners identified vowels with greater accuracy when the vowel-specific pattern of spectral change was preserved at the vowel's center. The results from these two lines of research suggest that neither the vowel target nor consonantal transitions alone are fully sufficient for vowel identification.…”

The role of f 0 and formant frequencies in distinguishing the voices of men and women

“…For example, an F 1 value of 750 Hz might suggest either / / (or perhaps / /) spoken by a man or / / spoken by a woman. This gives rise to a circularity that is seldom discussed: Pattern recognition studies have shown that vowels can be categorized with considerably greater accuracy with the inclusion of f 0 as a normalizing parameter (e.g., Disner, 1980;Hillenbrand & Gayvert, 1993;Hillenbrand et al, 1995;J. D. Miller, 1989;Nearey, 1978;Nearey, Hogan, & Rozsypal, 1979;Syrdal & Gopal, 1986).…”

Spectral Study with Automatic Formant Extraction to Improve Non-native Pronunciation of English Vowels

“…Formant frequencies have been already well-studied in both American and British English vowels [2][3][4][5][6][7]. On another note, remarkable numerical investigations were performed by Jan Awrejcewicz involving vocal cord oscillations and primary resonances [8,9] and other particular effects as stability and bifurcation phenomena [10].…”