Automatic soundscape classification via comparative psychometrics and machine learning

Krithika Rajagopal, Phil Minnick, Colby Leider

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Computational acoustical ecology is a relatively new field in which long-term environmental recordings are mined for meaningful data. Humans quite naturally and automatically associate environmental sounds with emotions and can easily identify the components of a soundscape. However, equipping a computer to accurately and automatically rate unknown environmental recordings along subjective psychoacoustic dimensions, let alone report the environment (e.g., beach, barnyard, home kitchen, research lab, etc.) in which the environmental recordings were made with a high degree of accuracy is quite difficult. We present here a robust algorithm for automatic soundscape classification in which both psychometric data and computed audio features are compared and used to train a Naive Bayesian classifier. An algorithm for classifying the type of soundscape across different categories was developed. In a pilot test, automatic classification accuracy of 88% was achieved on 20 soundscapes, and the classifier was able to outperform human ratings in some tests. In a second test, classification accuracy of 95% was achieved on 30 soundscapes.

Original languageEnglish (US)
Title of host publication131st Audio Engineering Society Convention 2011
Pages992-999
Number of pages8
Volume2
StatePublished - 2011
Event131st Audio Engineering Society Convention 2011 - New York, NY, United States
Duration: Oct 20 2011Oct 23 2011

Other

Other131st Audio Engineering Society Convention 2011
CountryUnited States
CityNew York, NY
Period10/20/1110/23/11

Fingerprint

psychometrics
machine learning
Psychometrics
learning
Learning systems
Machine Learning
recording
classifiers
Classifiers
test pilots
emotions
psychoacoustics
ecology
Kitchens
beaches
ratings
Ecology
Beaches
classifying
Bayesian Classifier

ASJC Scopus subject areas

  • Modeling and Simulation
  • Acoustics and Ultrasonics

Cite this

Rajagopal, K., Minnick, P., & Leider, C. (2011). Automatic soundscape classification via comparative psychometrics and machine learning. In 131st Audio Engineering Society Convention 2011 (Vol. 2, pp. 992-999)

Automatic soundscape classification via comparative psychometrics and machine learning. / Rajagopal, Krithika; Minnick, Phil; Leider, Colby.

131st Audio Engineering Society Convention 2011. Vol. 2 2011. p. 992-999.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rajagopal, K, Minnick, P & Leider, C 2011, Automatic soundscape classification via comparative psychometrics and machine learning. in 131st Audio Engineering Society Convention 2011. vol. 2, pp. 992-999, 131st Audio Engineering Society Convention 2011, New York, NY, United States, 10/20/11.
Rajagopal K, Minnick P, Leider C. Automatic soundscape classification via comparative psychometrics and machine learning. In 131st Audio Engineering Society Convention 2011. Vol. 2. 2011. p. 992-999
Rajagopal, Krithika ; Minnick, Phil ; Leider, Colby. / Automatic soundscape classification via comparative psychometrics and machine learning. 131st Audio Engineering Society Convention 2011. Vol. 2 2011. pp. 992-999
@inproceedings{687bc48809f04feca9e0eaef136320d0,
title = "Automatic soundscape classification via comparative psychometrics and machine learning",
abstract = "Computational acoustical ecology is a relatively new field in which long-term environmental recordings are mined for meaningful data. Humans quite naturally and automatically associate environmental sounds with emotions and can easily identify the components of a soundscape. However, equipping a computer to accurately and automatically rate unknown environmental recordings along subjective psychoacoustic dimensions, let alone report the environment (e.g., beach, barnyard, home kitchen, research lab, etc.) in which the environmental recordings were made with a high degree of accuracy is quite difficult. We present here a robust algorithm for automatic soundscape classification in which both psychometric data and computed audio features are compared and used to train a Naive Bayesian classifier. An algorithm for classifying the type of soundscape across different categories was developed. In a pilot test, automatic classification accuracy of 88{\%} was achieved on 20 soundscapes, and the classifier was able to outperform human ratings in some tests. In a second test, classification accuracy of 95{\%} was achieved on 30 soundscapes.",
author = "Krithika Rajagopal and Phil Minnick and Colby Leider",
year = "2011",
language = "English (US)",
isbn = "9781618393968",
volume = "2",
pages = "992--999",
booktitle = "131st Audio Engineering Society Convention 2011",

}

TY - GEN

T1 - Automatic soundscape classification via comparative psychometrics and machine learning

AU - Rajagopal, Krithika

AU - Minnick, Phil

AU - Leider, Colby

PY - 2011

Y1 - 2011

N2 - Computational acoustical ecology is a relatively new field in which long-term environmental recordings are mined for meaningful data. Humans quite naturally and automatically associate environmental sounds with emotions and can easily identify the components of a soundscape. However, equipping a computer to accurately and automatically rate unknown environmental recordings along subjective psychoacoustic dimensions, let alone report the environment (e.g., beach, barnyard, home kitchen, research lab, etc.) in which the environmental recordings were made with a high degree of accuracy is quite difficult. We present here a robust algorithm for automatic soundscape classification in which both psychometric data and computed audio features are compared and used to train a Naive Bayesian classifier. An algorithm for classifying the type of soundscape across different categories was developed. In a pilot test, automatic classification accuracy of 88% was achieved on 20 soundscapes, and the classifier was able to outperform human ratings in some tests. In a second test, classification accuracy of 95% was achieved on 30 soundscapes.

AB - Computational acoustical ecology is a relatively new field in which long-term environmental recordings are mined for meaningful data. Humans quite naturally and automatically associate environmental sounds with emotions and can easily identify the components of a soundscape. However, equipping a computer to accurately and automatically rate unknown environmental recordings along subjective psychoacoustic dimensions, let alone report the environment (e.g., beach, barnyard, home kitchen, research lab, etc.) in which the environmental recordings were made with a high degree of accuracy is quite difficult. We present here a robust algorithm for automatic soundscape classification in which both psychometric data and computed audio features are compared and used to train a Naive Bayesian classifier. An algorithm for classifying the type of soundscape across different categories was developed. In a pilot test, automatic classification accuracy of 88% was achieved on 20 soundscapes, and the classifier was able to outperform human ratings in some tests. In a second test, classification accuracy of 95% was achieved on 30 soundscapes.

UR - http://www.scopus.com/inward/record.url?scp=84866342101&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866342101&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84866342101

SN - 9781618393968

VL - 2

SP - 992

EP - 999

BT - 131st Audio Engineering Society Convention 2011

ER -