Enhanced LBP texture features from time frequency representations for acoustic scene classification

Shamsiah Abidin, Roberto Togneri, Ferdous Sohel

Research output: Chapter in Book/Conference paperConference paper

5 Citations (Scopus)

Abstract

This paper introduces the use of local binary patterns (LBP) extracted from a time-frequency representation (TFR) for acoustic scene classification. As LBP provides a description of the global TFR texture we propose a novel zoning mechanism that provides a simple solution to extract spectrally relevant local features which better characterize the audio TFRs. To further improve the classification performance, we perform feature and score level fusion of the proposed LBP (with zoning) with histogram of gradients (HOG) of the TFR images. Our technique demonstrates an improved performance by achieving a classification accuracy of 95.2% using a fusion of time-frequency derived features.

Original languageEnglish
Title of host publication2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings
EditorsMagdy A. Bayoumi
Place of PublicationNew Orleans
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages626-630
Number of pages5
ISBN (Electronic)9781509041176
DOIs
Publication statusPublished - 19 Jun 2017
Event2017 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2017 - New Orleans, United States
Duration: 5 Mar 20179 Mar 2017

Conference

Conference2017 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2017
CountryUnited States
CityNew Orleans
Period5/03/179/03/17

Fingerprint

Zoning
Textures
Acoustics
Fusion reactions

Cite this

Abidin, S., Togneri, R., & Sohel, F. (2017). Enhanced LBP texture features from time frequency representations for acoustic scene classification. In M. A. Bayoumi (Ed.), 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings (pp. 626-630). [7952231] New Orleans: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICASSP.2017.7952231
Abidin, Shamsiah ; Togneri, Roberto ; Sohel, Ferdous. / Enhanced LBP texture features from time frequency representations for acoustic scene classification. 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings. editor / Magdy A. Bayoumi. New Orleans : IEEE, Institute of Electrical and Electronics Engineers, 2017. pp. 626-630
@inproceedings{99cabb1282f041fb82fcfa2fc4912063,
title = "Enhanced LBP texture features from time frequency representations for acoustic scene classification",
abstract = "This paper introduces the use of local binary patterns (LBP) extracted from a time-frequency representation (TFR) for acoustic scene classification. As LBP provides a description of the global TFR texture we propose a novel zoning mechanism that provides a simple solution to extract spectrally relevant local features which better characterize the audio TFRs. To further improve the classification performance, we perform feature and score level fusion of the proposed LBP (with zoning) with histogram of gradients (HOG) of the TFR images. Our technique demonstrates an improved performance by achieving a classification accuracy of 95.2{\%} using a fusion of time-frequency derived features.",
keywords = "acoustic scene, feature extraction, fusion, local binary patterns, time-frequency analysis",
author = "Shamsiah Abidin and Roberto Togneri and Ferdous Sohel",
year = "2017",
month = "6",
day = "19",
doi = "10.1109/ICASSP.2017.7952231",
language = "English",
pages = "626--630",
editor = "Bayoumi, {Magdy A. }",
booktitle = "2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

Abidin, S, Togneri, R & Sohel, F 2017, Enhanced LBP texture features from time frequency representations for acoustic scene classification. in MA Bayoumi (ed.), 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings., 7952231, IEEE, Institute of Electrical and Electronics Engineers, New Orleans, pp. 626-630, 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2017, New Orleans, United States, 5/03/17. https://doi.org/10.1109/ICASSP.2017.7952231

Enhanced LBP texture features from time frequency representations for acoustic scene classification. / Abidin, Shamsiah; Togneri, Roberto; Sohel, Ferdous.

2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings. ed. / Magdy A. Bayoumi. New Orleans : IEEE, Institute of Electrical and Electronics Engineers, 2017. p. 626-630 7952231.

Research output: Chapter in Book/Conference paperConference paper

TY - GEN

T1 - Enhanced LBP texture features from time frequency representations for acoustic scene classification

AU - Abidin, Shamsiah

AU - Togneri, Roberto

AU - Sohel, Ferdous

PY - 2017/6/19

Y1 - 2017/6/19

N2 - This paper introduces the use of local binary patterns (LBP) extracted from a time-frequency representation (TFR) for acoustic scene classification. As LBP provides a description of the global TFR texture we propose a novel zoning mechanism that provides a simple solution to extract spectrally relevant local features which better characterize the audio TFRs. To further improve the classification performance, we perform feature and score level fusion of the proposed LBP (with zoning) with histogram of gradients (HOG) of the TFR images. Our technique demonstrates an improved performance by achieving a classification accuracy of 95.2% using a fusion of time-frequency derived features.

AB - This paper introduces the use of local binary patterns (LBP) extracted from a time-frequency representation (TFR) for acoustic scene classification. As LBP provides a description of the global TFR texture we propose a novel zoning mechanism that provides a simple solution to extract spectrally relevant local features which better characterize the audio TFRs. To further improve the classification performance, we perform feature and score level fusion of the proposed LBP (with zoning) with histogram of gradients (HOG) of the TFR images. Our technique demonstrates an improved performance by achieving a classification accuracy of 95.2% using a fusion of time-frequency derived features.

KW - acoustic scene

KW - feature extraction

KW - fusion

KW - local binary patterns

KW - time-frequency analysis

UR - http://www.scopus.com/inward/record.url?scp=85023757950&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2017.7952231

DO - 10.1109/ICASSP.2017.7952231

M3 - Conference paper

SP - 626

EP - 630

BT - 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings

A2 - Bayoumi, Magdy A.

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - New Orleans

ER -

Abidin S, Togneri R, Sohel F. Enhanced LBP texture features from time frequency representations for acoustic scene classification. In Bayoumi MA, editor, 2017 IEEE International Conference on Acoustics, Speech, and Signal Processing Proceedings. New Orleans: IEEE, Institute of Electrical and Electronics Engineers. 2017. p. 626-630. 7952231 https://doi.org/10.1109/ICASSP.2017.7952231