Automatic Gender Detection Based on Characteristics of Vocal Folds for Mobile Healthcare System

Musaed Alhussein, Zulfiqar Ali, Muhammad Imran, Wadood Abdul

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

An automatic gender detection may be useful in some cases of a mobile healthcare system. For example, there are some pathologies, such as vocal fold cyst, which mainly occur in female patients. If there is an automatic method for gender detection embedded into the system, it is easy for a healthcare professional to assess and prescribe appropriate medication to the patient. In human voice production system, contribution of the vocal folds is very vital. The length of the vocal folds is gender dependent; a male speaker has longer vocal folds than a female speaker. Due to longer vocal folds, the voice of a male becomes heavy and, therefore, contains more voice intensity. Based on this idea, a new type of time domain acoustic feature for automatic gender detection system is proposed in this paper. The proposed feature measures the voice intensity by calculating the area under the modified voice contour to make the differentiation between males and females. Two different databases are used to show that the proposed feature is independent of text, spoken language, dialect region, recording system, and environment. The obtained results for clean and noisy speech are 98.27% and 96.55%, respectively.

LanguageEnglish
Article number7805217
Pages1-12
Number of pages12
JournalMobile Information Systems
Volume2016
DOIs
Publication statusPublished - 2016

Fingerprint

Pathology
mHealth
Acoustics

Cite this

@article{1262deb3a812457f8b295f3db0461ec9,
title = "Automatic Gender Detection Based on Characteristics of Vocal Folds for Mobile Healthcare System",
abstract = "An automatic gender detection may be useful in some cases of a mobile healthcare system. For example, there are some pathologies, such as vocal fold cyst, which mainly occur in female patients. If there is an automatic method for gender detection embedded into the system, it is easy for a healthcare professional to assess and prescribe appropriate medication to the patient. In human voice production system, contribution of the vocal folds is very vital. The length of the vocal folds is gender dependent; a male speaker has longer vocal folds than a female speaker. Due to longer vocal folds, the voice of a male becomes heavy and, therefore, contains more voice intensity. Based on this idea, a new type of time domain acoustic feature for automatic gender detection system is proposed in this paper. The proposed feature measures the voice intensity by calculating the area under the modified voice contour to make the differentiation between males and females. Two different databases are used to show that the proposed feature is independent of text, spoken language, dialect region, recording system, and environment. The obtained results for clean and noisy speech are 98.27{\%} and 96.55{\%}, respectively.",
author = "Musaed Alhussein and Zulfiqar Ali and Muhammad Imran and Wadood Abdul",
year = "2016",
doi = "10.1155/2016/7805217",
language = "English",
volume = "2016",
pages = "1--12",
journal = "Mobile Information Systems",
issn = "1574-017X",

}

Automatic Gender Detection Based on Characteristics of Vocal Folds for Mobile Healthcare System. / Alhussein, Musaed; Ali, Zulfiqar; Imran, Muhammad; Abdul, Wadood.

In: Mobile Information Systems, Vol. 2016, 7805217, 2016, p. 1-12.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Automatic Gender Detection Based on Characteristics of Vocal Folds for Mobile Healthcare System

AU - Alhussein, Musaed

AU - Ali, Zulfiqar

AU - Imran, Muhammad

AU - Abdul, Wadood

PY - 2016

Y1 - 2016

N2 - An automatic gender detection may be useful in some cases of a mobile healthcare system. For example, there are some pathologies, such as vocal fold cyst, which mainly occur in female patients. If there is an automatic method for gender detection embedded into the system, it is easy for a healthcare professional to assess and prescribe appropriate medication to the patient. In human voice production system, contribution of the vocal folds is very vital. The length of the vocal folds is gender dependent; a male speaker has longer vocal folds than a female speaker. Due to longer vocal folds, the voice of a male becomes heavy and, therefore, contains more voice intensity. Based on this idea, a new type of time domain acoustic feature for automatic gender detection system is proposed in this paper. The proposed feature measures the voice intensity by calculating the area under the modified voice contour to make the differentiation between males and females. Two different databases are used to show that the proposed feature is independent of text, spoken language, dialect region, recording system, and environment. The obtained results for clean and noisy speech are 98.27% and 96.55%, respectively.

AB - An automatic gender detection may be useful in some cases of a mobile healthcare system. For example, there are some pathologies, such as vocal fold cyst, which mainly occur in female patients. If there is an automatic method for gender detection embedded into the system, it is easy for a healthcare professional to assess and prescribe appropriate medication to the patient. In human voice production system, contribution of the vocal folds is very vital. The length of the vocal folds is gender dependent; a male speaker has longer vocal folds than a female speaker. Due to longer vocal folds, the voice of a male becomes heavy and, therefore, contains more voice intensity. Based on this idea, a new type of time domain acoustic feature for automatic gender detection system is proposed in this paper. The proposed feature measures the voice intensity by calculating the area under the modified voice contour to make the differentiation between males and females. Two different databases are used to show that the proposed feature is independent of text, spoken language, dialect region, recording system, and environment. The obtained results for clean and noisy speech are 98.27% and 96.55%, respectively.

UR - http://www.scopus.com/inward/record.url?scp=84973444145&partnerID=8YFLogxK

U2 - 10.1155/2016/7805217

DO - 10.1155/2016/7805217

M3 - Article

VL - 2016

SP - 1

EP - 12

JO - Mobile Information Systems

T2 - Mobile Information Systems

JF - Mobile Information Systems

SN - 1574-017X

M1 - 7805217

ER -