Spoken term detection of zero-resource language using machine learning

Akinori Ito; Masatoshi Koizumi

doi:10.1145/3193063.3193068

Spoken term detection of zero-resource language using machine learning

Akinori Ito, Masatoshi Koizumi

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Citations (Scopus)

Abstract

In this paper, we propose a spoken term detection method for detection of terms in zero-resource languages. The proposed method uses the classifier (the speech comparator) trained by a machine learning method combined with the dynamic time warping method. The advantage of the proposed method is that the classifier can be trained using a large language resource that is different from the target language. We exploited the random forest as a classifier, and carried out an experiment of the spoken term detection from Kaqchikel speech. As a result, the proposed method showed better detection performance compared with the method based on the Euclidean distance.

Original language	English
Title of host publication	2018 International Conference on Intelligent Information Technology, ICIIT 2018
Publisher	Association for Computing Machinery
Pages	45-49
Number of pages	5
ISBN (Electronic)	9781450363785
DOIs	https://doi.org/10.1145/3193063.3193068
Publication status	Published - 2018 Feb 26
Event	2018 International Conference on Intelligent Information Technology, ICIIT 2018 - Hanoi, Viet Nam Duration: 2018 Feb 26 → 2018 Feb 28

Publication series

Name	ACM International Conference Proceeding Series

Conference

Conference	2018 International Conference on Intelligent Information Technology, ICIIT 2018
Country/Territory	Viet Nam
City	Hanoi
Period	18/2/26 → 18/2/28

Keywords

Dynamic time warping
Kaqchikel
Random forest
Spoken term detection

Access to Document

10.1145/3193063.3193068

Cite this

@inproceedings{2375c0957cfa4d97abd4cdbe109e4c97,

title = "Spoken term detection of zero-resource language using machine learning",

abstract = "In this paper, we propose a spoken term detection method for detection of terms in zero-resource languages. The proposed method uses the classifier (the speech comparator) trained by a machine learning method combined with the dynamic time warping method. The advantage of the proposed method is that the classifier can be trained using a large language resource that is different from the target language. We exploited the random forest as a classifier, and carried out an experiment of the spoken term detection from Kaqchikel speech. As a result, the proposed method showed better detection performance compared with the method based on the Euclidean distance.",

keywords = "Dynamic time warping, Kaqchikel, Random forest, Spoken term detection",

author = "Akinori Ito and Masatoshi Koizumi",

note = "Funding Information: This work was supported by Program for Key Interdisciplinary Research, Tohoku University. Publisher Copyright: {\textcopyright} 2018 Association for Computing Machinery.; 2018 International Conference on Intelligent Information Technology, ICIIT 2018 ; Conference date: 26-02-2018 Through 28-02-2018",

year = "2018",

month = feb,

day = "26",

doi = "10.1145/3193063.3193068",

language = "English",

series = "ACM International Conference Proceeding Series",

publisher = "Association for Computing Machinery",

pages = "45--49",

booktitle = "2018 International Conference on Intelligent Information Technology, ICIIT 2018",

}

Ito, A & Koizumi, M 2018, Spoken term detection of zero-resource language using machine learning. in 2018 International Conference on Intelligent Information Technology, ICIIT 2018. ACM International Conference Proceeding Series, Association for Computing Machinery, pp. 45-49, 2018 International Conference on Intelligent Information Technology, ICIIT 2018, Hanoi, Viet Nam, 18/2/26. https://doi.org/10.1145/3193063.3193068

Spoken term detection of zero-resource language using machine learning. / Ito, Akinori ; Koizumi, Masatoshi.
2018 International Conference on Intelligent Information Technology, ICIIT 2018. Association for Computing Machinery, 2018. p. 45-49 (ACM International Conference Proceeding Series).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Spoken term detection of zero-resource language using machine learning

AU - Ito, Akinori

AU - Koizumi, Masatoshi

PY - 2018/2/26

Y1 - 2018/2/26

N2 - In this paper, we propose a spoken term detection method for detection of terms in zero-resource languages. The proposed method uses the classifier (the speech comparator) trained by a machine learning method combined with the dynamic time warping method. The advantage of the proposed method is that the classifier can be trained using a large language resource that is different from the target language. We exploited the random forest as a classifier, and carried out an experiment of the spoken term detection from Kaqchikel speech. As a result, the proposed method showed better detection performance compared with the method based on the Euclidean distance.

AB - In this paper, we propose a spoken term detection method for detection of terms in zero-resource languages. The proposed method uses the classifier (the speech comparator) trained by a machine learning method combined with the dynamic time warping method. The advantage of the proposed method is that the classifier can be trained using a large language resource that is different from the target language. We exploited the random forest as a classifier, and carried out an experiment of the spoken term detection from Kaqchikel speech. As a result, the proposed method showed better detection performance compared with the method based on the Euclidean distance.

KW - Dynamic time warping

KW - Kaqchikel

KW - Random forest

KW - Spoken term detection

UR - http://www.scopus.com/inward/record.url?scp=85047252470&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047252470&partnerID=8YFLogxK

U2 - 10.1145/3193063.3193068

DO - 10.1145/3193063.3193068

M3 - Conference contribution

AN - SCOPUS:85047252470

T3 - ACM International Conference Proceeding Series

SP - 45

EP - 49

BT - 2018 International Conference on Intelligent Information Technology, ICIIT 2018

PB - Association for Computing Machinery

T2 - 2018 International Conference on Intelligent Information Technology, ICIIT 2018

Y2 - 26 February 2018 through 28 February 2018

ER -

Spoken term detection of zero-resource language using machine learning

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this