Automated classification of vowel category and speaker type in the high-frequency spectrum
The high-frequency region of vowel signals (above the third formant or F3) has received little research attention. Recent evidence, however, has documented the perceptual utility of high-frequency information in the speech signal above the traditional frequency bandwidth known to contain important c...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2016-04-01
|
Series: | Audiology Research |
Subjects: | |
Online Access: | https://audiologyresearch.org/index.php/audio/article/view/137 |
id |
doaj-fc96d44bf7e44df886642c46f64e086c |
---|---|
record_format |
Article |
spelling |
doaj-fc96d44bf7e44df886642c46f64e086c2021-01-02T10:53:06ZengMDPI AGAudiology Research2039-43302039-43492016-04-016110.4081/audiores.2016.13786Automated classification of vowel category and speaker type in the high-frequency spectrumJeremy J. Donai0Saeid Motiian1Gianfranco Doretto2Department of Communication Sciences and Disorders, West Virginia University, Morgantown, WVLane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WVLane Department of Computer Science and Electrical Engineering, West Virginia University, Morgantown, WVThe high-frequency region of vowel signals (above the third formant or F3) has received little research attention. Recent evidence, however, has documented the perceptual utility of high-frequency information in the speech signal above the traditional frequency bandwidth known to contain important cues for speech and speaker recognition. The purpose of this study was to determine if high-pass filtered vowels could be separated by vowel category and speaker type in a supervised learning framework. Mel frequency cepstral coefficients (MFCCs) were extracted from productions of six vowel categories produced by two male, two female, and two child speakers. Results revealed that the filtered vowels were well separated by vowel category and speaker type using MFCCs from the high-frequency spectrum. This demonstrates the presence of useful information for automated classification from the high-frequency region and is the first study to report findings of this nature in a supervised learning framework.https://audiologyresearch.org/index.php/audio/article/view/137Classificationformantshigh-frequencymel frequency cepstral coefficientsvowels. |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Jeremy J. Donai Saeid Motiian Gianfranco Doretto |
spellingShingle |
Jeremy J. Donai Saeid Motiian Gianfranco Doretto Automated classification of vowel category and speaker type in the high-frequency spectrum Audiology Research Classification formants high-frequency mel frequency cepstral coefficients vowels. |
author_facet |
Jeremy J. Donai Saeid Motiian Gianfranco Doretto |
author_sort |
Jeremy J. Donai |
title |
Automated classification of vowel category and speaker type in the high-frequency spectrum |
title_short |
Automated classification of vowel category and speaker type in the high-frequency spectrum |
title_full |
Automated classification of vowel category and speaker type in the high-frequency spectrum |
title_fullStr |
Automated classification of vowel category and speaker type in the high-frequency spectrum |
title_full_unstemmed |
Automated classification of vowel category and speaker type in the high-frequency spectrum |
title_sort |
automated classification of vowel category and speaker type in the high-frequency spectrum |
publisher |
MDPI AG |
series |
Audiology Research |
issn |
2039-4330 2039-4349 |
publishDate |
2016-04-01 |
description |
The high-frequency region of vowel signals (above the third formant or F3) has received little research attention. Recent evidence, however, has documented the perceptual utility of high-frequency information in the speech signal above the traditional frequency bandwidth known to contain important cues for speech and speaker recognition. The purpose of this study was to determine if high-pass filtered vowels could be separated by vowel category and speaker type in a supervised learning framework. Mel frequency cepstral coefficients (MFCCs) were extracted from productions of six vowel categories produced by two male, two female, and two child speakers. Results revealed that the filtered vowels were well separated by vowel category and speaker type using MFCCs from the high-frequency spectrum. This demonstrates the presence of useful information for automated classification from the high-frequency region and is the first study to report findings of this nature in a supervised learning framework. |
topic |
Classification formants high-frequency mel frequency cepstral coefficients vowels. |
url |
https://audiologyresearch.org/index.php/audio/article/view/137 |
work_keys_str_mv |
AT jeremyjdonai automatedclassificationofvowelcategoryandspeakertypeinthehighfrequencyspectrum AT saeidmotiian automatedclassificationofvowelcategoryandspeakertypeinthehighfrequencyspectrum AT gianfrancodoretto automatedclassificationofvowelcategoryandspeakertypeinthehighfrequencyspectrum |
_version_ |
1724355358936268800 |