An improved speech/nonspeech classification based on feature combination for audio indexing

Ji Soo Keum; Hyon Soo Lee; Masafumi Hagiwara

doi:10.1587/transfun.E93.A.830

An improved speech/nonspeech classification based on feature combination for audio indexing

Ji Soo Keum, Hyon Soo Lee, Masafumi Hagiwara

Department of Information and Computer Science

Research output: Contribution to journal › Article › peer-review

Abstract

In this letter, we propose an improved speech/nonspeech classification method to effectively classify a multimedia source. To improve performance, we introduce a feature based on spectral duration analysis, and combine recently proposed features such as high zero crossing rate ratio (HZCRR), low short time energy ratio (LSTER), and pitch ratio (PR). According to the results of our experiments on speech, music, and environmental sounds, the proposed method obtained high classification results when compared with conventional approaches.

Original language	English
Pages (from-to)	830-832
Number of pages	3
Journal	IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
Volume	E93-A
Issue number	4
DOIs	https://doi.org/10.1587/transfun.E93.A.830
Publication status	Published - 2010 Apr

Keywords

Audio indexing
Feature combination
Spectral duration analysis
Speech/nonspeech classification

ASJC Scopus subject areas

Signal Processing
Computer Graphics and Computer-Aided Design
Electrical and Electronic Engineering
Applied Mathematics

Access to Document

10.1587/transfun.E93.A.830

Cite this

@article{956b68801374459bad511027a4228a16,

title = "An improved speech/nonspeech classification based on feature combination for audio indexing",

abstract = "In this letter, we propose an improved speech/nonspeech classification method to effectively classify a multimedia source. To improve performance, we introduce a feature based on spectral duration analysis, and combine recently proposed features such as high zero crossing rate ratio (HZCRR), low short time energy ratio (LSTER), and pitch ratio (PR). According to the results of our experiments on speech, music, and environmental sounds, the proposed method obtained high classification results when compared with conventional approaches.",

keywords = "Audio indexing, Feature combination, Spectral duration analysis, Speech/nonspeech classification",

author = "Keum, {Ji Soo} and Lee, {Hyon Soo} and Masafumi Hagiwara",

year = "2010",

month = apr,

doi = "10.1587/transfun.E93.A.830",

language = "English",

volume = "E93-A",

pages = "830--832",

journal = "IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences",

issn = "0916-8508",

publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",

number = "4",

}

TY - JOUR

T1 - An improved speech/nonspeech classification based on feature combination for audio indexing

AU - Keum, Ji Soo

AU - Lee, Hyon Soo

AU - Hagiwara, Masafumi

PY - 2010/4

Y1 - 2010/4

N2 - In this letter, we propose an improved speech/nonspeech classification method to effectively classify a multimedia source. To improve performance, we introduce a feature based on spectral duration analysis, and combine recently proposed features such as high zero crossing rate ratio (HZCRR), low short time energy ratio (LSTER), and pitch ratio (PR). According to the results of our experiments on speech, music, and environmental sounds, the proposed method obtained high classification results when compared with conventional approaches.

AB - In this letter, we propose an improved speech/nonspeech classification method to effectively classify a multimedia source. To improve performance, we introduce a feature based on spectral duration analysis, and combine recently proposed features such as high zero crossing rate ratio (HZCRR), low short time energy ratio (LSTER), and pitch ratio (PR). According to the results of our experiments on speech, music, and environmental sounds, the proposed method obtained high classification results when compared with conventional approaches.

KW - Audio indexing

KW - Feature combination

KW - Spectral duration analysis

KW - Speech/nonspeech classification

UR - http://www.scopus.com/inward/record.url?scp=77950796943&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77950796943&partnerID=8YFLogxK

U2 - 10.1587/transfun.E93.A.830

DO - 10.1587/transfun.E93.A.830

M3 - Article

AN - SCOPUS:77950796943

SN - 0916-8508

VL - E93-A

SP - 830

EP - 832

JO - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

JF - IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences

IS - 4

ER -

An improved speech/nonspeech classification based on feature combination for audio indexing

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this