Classification of dog barks: A machine learning approach

Csaba Molnár, Frédéric Kaplan, Pierre Roy, François Pachet, P. Pongrácz, A. Dóka, A. Miklósi

Research output: Contribution to journalArticle

35 Citations (Scopus)

Abstract

In this study we analyzed the possible context-specific and individual-specific features of dog barks using a new machine-learning algorithm. A pool containing more than 6,000 barks, which were recorded in six different communicative situations was used as the sound sample. The algorithm's task was to learn which acoustic features of the barks, which were recorded in different contexts and from different individuals, could be distinguished from another. The program conducted this task by analyzing barks emitted in previously identified contexts by identified dogs. After the best feature set had been obtained (with which the highest identification rate was achieved), the efficiency of the algorithm was tested in a classification task in which unknown barks were analyzed. The recognition rates we found were highly above chance level: the algorithm could categorize the barks according to their recorded situation with an efficiency of 43% and with an efficiency of 52% of the barking individuals. These findings suggest that dog barks have context-specific and individual-specific acoustic features. In our opinion, this machine learning method may provide an efficient tool for analyzing acoustic data in various behavioral studies.

Original languageEnglish
Pages (from-to)389-400
Number of pages12
JournalAnimal Cognition
Volume11
Issue number3
DOIs
Publication statusPublished - Jul 2008

Fingerprint

artificial intelligence
bark
Acoustics
Dogs
dogs
acoustics
acoustic data
dog
machine learning
Machine Learning

Keywords

  • Acoustic communication
  • Dog barks
  • Genetic programming
  • Machine learning

ASJC Scopus subject areas

  • Animal Science and Zoology
  • Behavioral Neuroscience

Cite this

Classification of dog barks : A machine learning approach. / Molnár, Csaba; Kaplan, Frédéric; Roy, Pierre; Pachet, François; Pongrácz, P.; Dóka, A.; Miklósi, A.

In: Animal Cognition, Vol. 11, No. 3, 07.2008, p. 389-400.

Research output: Contribution to journalArticle

Molnár, Csaba ; Kaplan, Frédéric ; Roy, Pierre ; Pachet, François ; Pongrácz, P. ; Dóka, A. ; Miklósi, A. / Classification of dog barks : A machine learning approach. In: Animal Cognition. 2008 ; Vol. 11, No. 3. pp. 389-400.
@article{d53a3fe090af4a12a802244178f27b14,
title = "Classification of dog barks: A machine learning approach",
abstract = "In this study we analyzed the possible context-specific and individual-specific features of dog barks using a new machine-learning algorithm. A pool containing more than 6,000 barks, which were recorded in six different communicative situations was used as the sound sample. The algorithm's task was to learn which acoustic features of the barks, which were recorded in different contexts and from different individuals, could be distinguished from another. The program conducted this task by analyzing barks emitted in previously identified contexts by identified dogs. After the best feature set had been obtained (with which the highest identification rate was achieved), the efficiency of the algorithm was tested in a classification task in which unknown barks were analyzed. The recognition rates we found were highly above chance level: the algorithm could categorize the barks according to their recorded situation with an efficiency of 43{\%} and with an efficiency of 52{\%} of the barking individuals. These findings suggest that dog barks have context-specific and individual-specific acoustic features. In our opinion, this machine learning method may provide an efficient tool for analyzing acoustic data in various behavioral studies.",
keywords = "Acoustic communication, Dog barks, Genetic programming, Machine learning",
author = "Csaba Moln{\'a}r and Fr{\'e}d{\'e}ric Kaplan and Pierre Roy and Fran{\cc}ois Pachet and P. Pongr{\'a}cz and A. D{\'o}ka and A. Mikl{\'o}si",
year = "2008",
month = "7",
doi = "10.1007/s10071-007-0129-9",
language = "English",
volume = "11",
pages = "389--400",
journal = "Animal Cognition",
issn = "1435-9448",
publisher = "Springer Verlag",
number = "3",

}

TY - JOUR

T1 - Classification of dog barks

T2 - A machine learning approach

AU - Molnár, Csaba

AU - Kaplan, Frédéric

AU - Roy, Pierre

AU - Pachet, François

AU - Pongrácz, P.

AU - Dóka, A.

AU - Miklósi, A.

PY - 2008/7

Y1 - 2008/7

N2 - In this study we analyzed the possible context-specific and individual-specific features of dog barks using a new machine-learning algorithm. A pool containing more than 6,000 barks, which were recorded in six different communicative situations was used as the sound sample. The algorithm's task was to learn which acoustic features of the barks, which were recorded in different contexts and from different individuals, could be distinguished from another. The program conducted this task by analyzing barks emitted in previously identified contexts by identified dogs. After the best feature set had been obtained (with which the highest identification rate was achieved), the efficiency of the algorithm was tested in a classification task in which unknown barks were analyzed. The recognition rates we found were highly above chance level: the algorithm could categorize the barks according to their recorded situation with an efficiency of 43% and with an efficiency of 52% of the barking individuals. These findings suggest that dog barks have context-specific and individual-specific acoustic features. In our opinion, this machine learning method may provide an efficient tool for analyzing acoustic data in various behavioral studies.

AB - In this study we analyzed the possible context-specific and individual-specific features of dog barks using a new machine-learning algorithm. A pool containing more than 6,000 barks, which were recorded in six different communicative situations was used as the sound sample. The algorithm's task was to learn which acoustic features of the barks, which were recorded in different contexts and from different individuals, could be distinguished from another. The program conducted this task by analyzing barks emitted in previously identified contexts by identified dogs. After the best feature set had been obtained (with which the highest identification rate was achieved), the efficiency of the algorithm was tested in a classification task in which unknown barks were analyzed. The recognition rates we found were highly above chance level: the algorithm could categorize the barks according to their recorded situation with an efficiency of 43% and with an efficiency of 52% of the barking individuals. These findings suggest that dog barks have context-specific and individual-specific acoustic features. In our opinion, this machine learning method may provide an efficient tool for analyzing acoustic data in various behavioral studies.

KW - Acoustic communication

KW - Dog barks

KW - Genetic programming

KW - Machine learning

UR - http://www.scopus.com/inward/record.url?scp=46549090030&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46549090030&partnerID=8YFLogxK

U2 - 10.1007/s10071-007-0129-9

DO - 10.1007/s10071-007-0129-9

M3 - Article

C2 - 18197442

AN - SCOPUS:46549090030

VL - 11

SP - 389

EP - 400

JO - Animal Cognition

JF - Animal Cognition

SN - 1435-9448

IS - 3

ER -