Classification of stop consonant place of articulation

Abstract

One of the approaches to automatic speech recognition is a distinctive feature-based speech recognition system, in which each of the underlying word segments is represented with a set of distinctive features. This thesis presents a study concerning acoustic attributes used for identifying the place of articulation features for stop consonant segments. The acoustic attributes are selected so that they capture the information relevant to place identification, including amplitude and energy of release bursts, formant movements of adjacent vowels, spectra of noises after the releases, and some temporal cues.An experimental procedure for examining the relative importance of these acoustic attributes for identifying stop place is developed. The ability of each attribute to separate the three places is evaluated by the classification error based on the distributions of its values for the three places, and another quantifier based on F-ratio. These two quantifiers generally agree and show how well each individual attribute separates the three places.Combinations of non-redundant attributes are used for the place classifications based on Mahalanobis distance. When stops contain release bursts, the classification accuracies are better than 90%. It was also shown that voicing and vowel frontness contexts lead to a better classification accuracy of stops in some contexts. When stops are located between two vowels, information on the formant structures in the vowels on both sides can be combined. Such combination yielded the best classification accuracy of 95.5%. By using appropriate methods for stops in different contexts, an overall classification accuracy of 92.1% is achieved.Linear discriminant function analysis is used to address the relative importance of these attributes when combinations are used. Their discriminating abilities and the ranking of their relative importance to the classifications in different vowel and voicing contexts are reported. The overall findings are that attributes relating to the burst spectrum in relation to the vowel contribute most effectively, while attributes relating to formant transition are somewhat less effective. The approach used in this study can be applied to different classes of sounds, as well as stops in different noise environments.

Classification of stop place in consonant-vowel contexts using feature extrapolation of acoustic-phonetic features in telephone speech

“…5,10 Av-Ahi is found as the log of the ratio of the amplitude of the first formant prominence measured at voicing onset to the maximum amplitude of the spectrum in the range from 3.5 to 8 kHz at burst release. 7,10 To obtain Ehi-E23 and Av-Ahi from narrowband speech, extrapolation techniques will be required.…”

“…Ehi-E23 has been used to represent the degree of the spectral tilt by comparing the energy of the spectrum between the high frequency band and the mid frequency band. 7 Ehi-E23 is found as the log of the ratio between the total energy of the burst spectrum in the range from 3.5 to 8 kHz and from 1.25 to 3 kHz. Av-Ahi represents the strength of the high frequency component of the burst release, compared to the amplitude of the first formant prominence of the adjacent vowel.…”

“…[7][8][9][10]33,34 Most cues for identifying stop consonants require spectral information, but in telephone speech, spectral information may not exist over 3.5 kHz, so that there are limitations in selecting appropriate acoustic cues.…”

“…2,3 Acoustic cues extracted in this approach are described as reducing speaker-dependent information and focusing on the linguistic information in the speech signal. [4][5][6] The distinctive feature-based approach has been examined in stop consonant classification, [7][8][9][10] where the stops include /p/, /t/, /k/, /b/, /d/, and /g/, in English. Two features are needed to classify each stop.…”

¿Lengua o hablante? – Investigando las alternancias coronal-velares en vilela

“…Dado que hay medidas acústicas muy robustas y de fácil acceso, a saber, la cantidad de energía del sonido y la distribución de esa energía, además de la duración hasta el inicio periódico de la vocal siguiente, esto sirve, a nuestro entender, para caracterizar la explosión de la oclusiva y los indicios incrustados respecto del punto de articulación. Esta es la razón por la que hemos elegido como parámetros el tiempo de inicio de la sonoridad (VOT), la intensidad basada en RMS (valor cuadrático medio) en el período entre la emisión y el inicio de la sonoridad y Amid-Ahi (Suchato 2004). Amid-Ahi es una medida espectral relativa del espectro de explosión que compara las amplitudes en las secciones medias (1250-3000Hz) y altas (3500-8000Hz) del espectro seleccionado.…”