Automatic Smoker Detection from Telephone Speech Signals

Amir Hossein Poorjam, Soheila Hesaraki, Saeid Safavi, Hugo Van hamme, Mohamad Hasan Bahari

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

1 Citation (Scopus)
71 Downloads (Pure)

Abstract

This paper proposes an automatic smoking habit detection from spontaneous telephone speech signals. In this method, each utterance is modeled using i-vector and non-negative factor analysis (NFA) frameworks, which yield low-dimensional representation of utterances by applying factor analysis on Gaussian mixture model means and weights respectively. Each framework is evaluated using different classification algorithms to detect the smoker speakers. Finally, score-level fusion of the i-vector-based and the NFA-based recognizers is considered to improve the classification accuracy. The proposed method is evaluated on telephone speech signals of speakers whose smoking habits are known drawn from the National Institute of Standards and Technology (NIST) 2008 and 2010 Speaker Recognition Evaluation databases. Experimental results over 1194 utterances show the effectiveness of the proposed approach for the automatic smoking habit detection task.
Original languageEnglish
Title of host publicationSpeech and Computer : 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings
EditorsAlexey Karpov, Rodmonga Potapova, Iosif Mporas
Number of pages11
PublisherSpringer
Publication date13 Aug 2017
Pages200-210
ISBN (Print)978-3-319-66428-6
ISBN (Electronic)978-3-319-66429-3
DOIs
Publication statusPublished - 13 Aug 2017
Event19th International Conference on Speech and Computer SPECOM 2017 - Hatfield, United Kingdom
Duration: 12 Sep 201716 Sep 2017

Conference

Conference19th International Conference on Speech and Computer SPECOM 2017
CountryUnited Kingdom
CityHatfield
Period12/09/201716/09/2017
SeriesLecture Notes in Computer Science
Volume10458
ISSN0302-9743

Fingerprint

Factor analysis
Telephone
Fusion reactions

Keywords

  • Smoker detection
  • i-Vector
  • Non-negative factor analysis
  • Score fusion
  • Logistic regression

Cite this

Poorjam, A. H., Hesaraki, S., Safavi, S., Van hamme, H., & Bahari, M. H. (2017). Automatic Smoker Detection from Telephone Speech Signals. In A. Karpov, R. Potapova, & I. Mporas (Eds.), Speech and Computer: 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings (pp. 200-210). Springer. Lecture Notes in Computer Science, Vol.. 10458 https://doi.org/10.1007/978-3-319-66429-3_19
Poorjam, Amir Hossein ; Hesaraki, Soheila ; Safavi, Saeid ; Van hamme, Hugo ; Bahari, Mohamad Hasan. / Automatic Smoker Detection from Telephone Speech Signals. Speech and Computer: 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings. editor / Alexey Karpov ; Rodmonga Potapova ; Iosif Mporas. Springer, 2017. pp. 200-210 (Lecture Notes in Computer Science, Vol. 10458).
@inproceedings{09b7d5bbcdea487db4c78ec00c41e3fc,
title = "Automatic Smoker Detection from Telephone Speech Signals",
abstract = "This paper proposes an automatic smoking habit detection from spontaneous telephone speech signals. In this method, each utterance is modeled using i-vector and non-negative factor analysis (NFA) frameworks, which yield low-dimensional representation of utterances by applying factor analysis on Gaussian mixture model means and weights respectively. Each framework is evaluated using different classification algorithms to detect the smoker speakers. Finally, score-level fusion of the i-vector-based and the NFA-based recognizers is considered to improve the classification accuracy. The proposed method is evaluated on telephone speech signals of speakers whose smoking habits are known drawn from the National Institute of Standards and Technology (NIST) 2008 and 2010 Speaker Recognition Evaluation databases. Experimental results over 1194 utterances show the effectiveness of the proposed approach for the automatic smoking habit detection task.",
keywords = "Smoker detection , i-Vector , Non-negative factor analysis, Score fusion, Logistic regression",
author = "Poorjam, {Amir Hossein} and Soheila Hesaraki and Saeid Safavi and {Van hamme}, Hugo and Bahari, {Mohamad Hasan}",
year = "2017",
month = "8",
day = "13",
doi = "10.1007/978-3-319-66429-3_19",
language = "English",
isbn = "978-3-319-66428-6",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "200--210",
editor = "Alexey Karpov and Rodmonga Potapova and Iosif Mporas",
booktitle = "Speech and Computer",
address = "Germany",

}

Poorjam, AH, Hesaraki, S, Safavi, S, Van hamme, H & Bahari, MH 2017, Automatic Smoker Detection from Telephone Speech Signals. in A Karpov, R Potapova & I Mporas (eds), Speech and Computer: 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings. Springer, Lecture Notes in Computer Science, vol. 10458, pp. 200-210, 19th International Conference on Speech and Computer SPECOM 2017, Hatfield, United Kingdom, 12/09/2017. https://doi.org/10.1007/978-3-319-66429-3_19

Automatic Smoker Detection from Telephone Speech Signals. / Poorjam, Amir Hossein; Hesaraki, Soheila; Safavi, Saeid; Van hamme, Hugo; Bahari, Mohamad Hasan.

Speech and Computer: 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings. ed. / Alexey Karpov; Rodmonga Potapova; Iosif Mporas. Springer, 2017. p. 200-210 (Lecture Notes in Computer Science, Vol. 10458).

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Automatic Smoker Detection from Telephone Speech Signals

AU - Poorjam, Amir Hossein

AU - Hesaraki, Soheila

AU - Safavi, Saeid

AU - Van hamme, Hugo

AU - Bahari, Mohamad Hasan

PY - 2017/8/13

Y1 - 2017/8/13

N2 - This paper proposes an automatic smoking habit detection from spontaneous telephone speech signals. In this method, each utterance is modeled using i-vector and non-negative factor analysis (NFA) frameworks, which yield low-dimensional representation of utterances by applying factor analysis on Gaussian mixture model means and weights respectively. Each framework is evaluated using different classification algorithms to detect the smoker speakers. Finally, score-level fusion of the i-vector-based and the NFA-based recognizers is considered to improve the classification accuracy. The proposed method is evaluated on telephone speech signals of speakers whose smoking habits are known drawn from the National Institute of Standards and Technology (NIST) 2008 and 2010 Speaker Recognition Evaluation databases. Experimental results over 1194 utterances show the effectiveness of the proposed approach for the automatic smoking habit detection task.

AB - This paper proposes an automatic smoking habit detection from spontaneous telephone speech signals. In this method, each utterance is modeled using i-vector and non-negative factor analysis (NFA) frameworks, which yield low-dimensional representation of utterances by applying factor analysis on Gaussian mixture model means and weights respectively. Each framework is evaluated using different classification algorithms to detect the smoker speakers. Finally, score-level fusion of the i-vector-based and the NFA-based recognizers is considered to improve the classification accuracy. The proposed method is evaluated on telephone speech signals of speakers whose smoking habits are known drawn from the National Institute of Standards and Technology (NIST) 2008 and 2010 Speaker Recognition Evaluation databases. Experimental results over 1194 utterances show the effectiveness of the proposed approach for the automatic smoking habit detection task.

KW - Smoker detection

KW - i-Vector

KW - Non-negative factor analysis

KW - Score fusion

KW - Logistic regression

U2 - 10.1007/978-3-319-66429-3_19

DO - 10.1007/978-3-319-66429-3_19

M3 - Article in proceeding

SN - 978-3-319-66428-6

T3 - Lecture Notes in Computer Science

SP - 200

EP - 210

BT - Speech and Computer

A2 - Karpov, Alexey

A2 - Potapova, Rodmonga

A2 - Mporas, Iosif

PB - Springer

ER -

Poorjam AH, Hesaraki S, Safavi S, Van hamme H, Bahari MH. Automatic Smoker Detection from Telephone Speech Signals. In Karpov A, Potapova R, Mporas I, editors, Speech and Computer: 19th International Conference, SPECOM 2017, Hatfield, UK, September 12-16, 2017, Proceedings. Springer. 2017. p. 200-210. (Lecture Notes in Computer Science, Vol. 10458). https://doi.org/10.1007/978-3-319-66429-3_19