Parametric representations of bird sounds for automatic species recognition

Panu Somervuo; Aki Härmä; Seppo Fagerlund

doi:10.1109/TASL.2006.872624

Parametric representations of bird sounds for automatic species recognition

Panu Somervuo^*
, Aki Härmä
, Seppo Fagerlund

^*Corresponding author for this work

Research output: Contribution to journal › Article › Academic › peer-review

Abstract

This paper is related to the development of signal processing techniques for automatic recognition of bird species. Three different parametric representations are compared. The first representation is based on sinusoidal modeling which has been earlier found useful for highly tonal bird sounds. Mel-cepstrum parameters are used since they have been found very useful in the parallel problem of speech recognition. Finally, a vector of various descriptive features is tested because such models are popular in audio classification applications, and bird song is almost like music. We briefly introduce the methods and evaluate their performance in the classification and recognition of both individual syllables and song fragments of 14 common North-European Passerine bird species

Original language	English
Article number	1709912
Pages (from-to)	2252-2263
Number of pages	12
Journal	IEEE Transactions on Audio, Speech and Language Processing
Volume	14
Issue number	6
DOIs	https://doi.org/10.1109/TASL.2006.872624
Publication status	Published - Nov 2006
Externally published	Yes

Keywords

Bird song
Dynamic time warping (DTW)
Feature extraction
Gaussian mixture model (GMM)
Hidden Markov model (HMM)
Sinusoidal modeling

Access to Document

10.1109/TASL.2006.872624

Cite this

@article{7ddbb854938846e58086eacb72c2cb3f,

title = "Parametric representations of bird sounds for automatic species recognition",

abstract = "This paper is related to the development of signal processing techniques for automatic recognition of bird species. Three different parametric representations are compared. The first representation is based on sinusoidal modeling which has been earlier found useful for highly tonal bird sounds. Mel-cepstrum parameters are used since they have been found very useful in the parallel problem of speech recognition. Finally, a vector of various descriptive features is tested because such models are popular in audio classification applications, and bird song is almost like music. We briefly introduce the methods and evaluate their performance in the classification and recognition of both individual syllables and song fragments of 14 common North-European Passerine bird species",

keywords = "Bird song, Dynamic time warping (DTW), Feature extraction, Gaussian mixture model (GMM), Hidden Markov model (HMM), Sinusoidal modeling",

author = "Panu Somervuo and Aki H{\"a}rm{\"a} and Seppo Fagerlund",

note = "Funding Information: Manuscript received February 2, 2005; revised November 1, 2005. The work of P. Somervuo was supported by the Academy of Finland Project 44886 (Finnish Centre of Excellence Programme 2000{\textendash}2005). The work of S. Fagerlund was supported by the Academy of Finland (The AveSound Project). The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Gerald Schuller.",

year = "2006",

month = nov,

doi = "10.1109/TASL.2006.872624",

language = "English",

volume = "14",

pages = "2252--2263",

journal = "IEEE Transactions on Audio, Speech and Language Processing",

issn = "1558-7916",

publisher = "IEEE",

number = "6",

}

TY - JOUR

T1 - Parametric representations of bird sounds for automatic species recognition

AU - Somervuo, Panu

AU - Härmä, Aki

AU - Fagerlund, Seppo

N1 - Funding Information: Manuscript received February 2, 2005; revised November 1, 2005. The work of P. Somervuo was supported by the Academy of Finland Project 44886 (Finnish Centre of Excellence Programme 2000–2005). The work of S. Fagerlund was supported by the Academy of Finland (The AveSound Project). The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Gerald Schuller.

PY - 2006/11

Y1 - 2006/11

N2 - This paper is related to the development of signal processing techniques for automatic recognition of bird species. Three different parametric representations are compared. The first representation is based on sinusoidal modeling which has been earlier found useful for highly tonal bird sounds. Mel-cepstrum parameters are used since they have been found very useful in the parallel problem of speech recognition. Finally, a vector of various descriptive features is tested because such models are popular in audio classification applications, and bird song is almost like music. We briefly introduce the methods and evaluate their performance in the classification and recognition of both individual syllables and song fragments of 14 common North-European Passerine bird species

AB - This paper is related to the development of signal processing techniques for automatic recognition of bird species. Three different parametric representations are compared. The first representation is based on sinusoidal modeling which has been earlier found useful for highly tonal bird sounds. Mel-cepstrum parameters are used since they have been found very useful in the parallel problem of speech recognition. Finally, a vector of various descriptive features is tested because such models are popular in audio classification applications, and bird song is almost like music. We briefly introduce the methods and evaluate their performance in the classification and recognition of both individual syllables and song fragments of 14 common North-European Passerine bird species

KW - Bird song

KW - Dynamic time warping (DTW)

KW - Feature extraction

KW - Gaussian mixture model (GMM)

KW - Hidden Markov model (HMM)

KW - Sinusoidal modeling

UR - https://www.scopus.com/pages/publications/34347345718

U2 - 10.1109/TASL.2006.872624

DO - 10.1109/TASL.2006.872624

M3 - Article

AN - SCOPUS:34347345718

SN - 1558-7916

VL - 14

SP - 2252

EP - 2263

JO - IEEE Transactions on Audio, Speech and Language Processing

JF - IEEE Transactions on Audio, Speech and Language Processing

IS - 6

M1 - 1709912

ER -

Parametric representations of bird sounds for automatic species recognition

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this