|
Avesound
|
Automatic recognition of bird species by their sounds
|
| Activity level |
A research project is looking for funding
|
| News |
| Introduction |
In Avesound project techniques and systems for automatic identification
of sounds produced by birds (Aves) are developed. The main emphasis of
Avesound is to develop algorithms for the automatic recognition of bird
species.
The long term goal of the research is to develop a system capable of
recognizing a majority of Finnish bird species from a continuous
environmental recording. The methods will be tested with species where
data from multiple individuals is already available or will be collected
during the project. Due to specific properties of bird sounds, it is
necessary to develop new signal processing and pattern recognition
techniques which are particularly suitable for the identification of
bird sounds. The approach is to take into account the physics of sound
production in birds and structural and statistical characteristics of
bird songs and calls. The developed methods have some important
applications in biology. At a higher level, a generic framework for
identification of animal sounds will be developed in the project.
This framework will be applicable also to the development of
identification techniques for other classes of animals such as mammals,
amphibians, and insects.
|
| Participants at HUT/Acoustics |
|
| Participants in other places |
Prof. Juha Tanttu and Jari Turunen (
TUT/Pori )
Mikko Ojanen ( UT/SERI)
Prof. Juhani Lokki ( UH/FMNH )
and many other more or less active collaborators.
|
| Publications |
- A. Härmä, ``Automatic recognition of bird species based on sinusoidal modeling
of syllables,'' in IEEE Int. Conf. Acoustics, Speech, Signal Processing
(ICASSP 2003), (Hong Kong), April 2003.pdf-file
- P . Somervuo and
A. Härmä, ``Analyzing bird song syllables on the
self-organizing map,'' in Proc. of Workshop on Self-Organizing
Maps (WSOM '03), (Hibikino, Japan), September 2003.pdf-file
- J.T. Tanttu, J. Turunen, and M. Ojanen,
"Automatic Classification of Flight Calls of Crossbill Species (Loxia
sp.)", 1st Conference on Acoustic Communication by Animals ,
(Maryland, USA), July 2003.pdf-file
- A. Härmä and P. Somervuo, CLASSIFICATION OF THE HARMONIC
STRUCTURE IN BIRD VOCALIZATION, in IEEE Int. Conf. Acoustics,
Speech, Signal Processing (ICASSP 2004), (Montreal,Canada), 17-21
May, 2004.pdf-file
- P. Somervuo and A. Härmä, BIRD SONG RECOGNITION BASED ON
SYLLABLE PAIR HISTOGRAMS, in IEEE Int. Conf. Acoustics,
Speech, Signal Processing (ICASSP 2004), (Montreal,Canada), 17-21
May, 2004.pdf-file
- A. Härmä, Avesound-memo: Phylloscopus-suvun lintujen
elementtien vertailu (In Finnish), Espoo, Finland, March,
2004. pdf-file
- S. Fagerlund, ``Acoustics and
physical models of bird sounds '' in Seminar in acoustics, HUT,
Laboratory of Acoustics and Audio Signal Processing, , (Espoo),
April 2004.pdf-file
- S. Fagerlund, ``Automatic
Recognition of Bird Species by Their Sounds '' Masters Thesis,
HUT, Laboratory of Acoustics and Audio Signal Processing, ,
(Espoo), 2004.thesis pdf-file ,
presentation pdf-file
- A. Selin, ``Bird sound
classification using wavelets'' Masters Thesis, TUT, Pori, ,
(Pori), 2005. thesis pdf-file
- S. Fagerlund and A. Härmä,
``Parametrization of inharmonic bird sounds for automatic
recognition'' 13th European Signal Processing Conference (EUSIPCO
2005), (Antalya, Turkey), 4-8 September, 2005. pdf-file
- A. Selin, J. Turunen and J. T. Tanttu,
``Bird sound recognition and classification using wavelets'' XX.
International BioAcoustics Congress (IBAC 2005), (Piran,
Slovenia), 15-19 September, 2005. abstract
- J. Turunen, A. Selin, J. T. Tanttu
and T. Lipping, ``De-noising aspects in the context of feature
extraction in automatic bird call recognition'' XX. International
BioAcoustics Congress (IBAC 2005), (Piran, Slovenia), 15-19
September, 2005. abstract
- J. T. Tanttu, J. Turunen, A. Selin and
M. Ojanen, ``Automatic Feature Extraction and Classification of
Crossbill (Loxia spp.) Flight Calls'' Bioacoustics,
Vol. 15, no. 3, pp. 251-269, 2006.
- A. Selin, J. Turunen and J. T. Tanttu,
``Wavelets in Automatic Recognition of Bird Sound'' EURASIP
Journal on Signal Processing Special Issue on Multirate Systems and
Applications, Vol 2007, Article ID 51806, 9 pages, 2006
- P. Somervuo, A. Härmä and
S. Fagerlund, ``Parametric representations of bird sounds for
automatic species recognition'' IEEE Trans. Speech and Audio
Processing, Vol. 14, no. 6, pp. 2252 - 2263, 2006.
- S. Fagerlund, ``Bird Species
Recognition Using Support Vector Machines'' accepted for
publication in EURASIP Journal on Applied Signal Processing (EURASIP
JASP) , 2007.
|
| Media (in Finnish mainly) |
-
Satakunnan Radio, Anneli Korpelainen, ``Satakunan Radion Uutiset'', 18.11.-03. mp3-file (0:31)
-
Satakunnan Radio, Pertti Kalinainen, ``Satakunan Luontoradio'', 18.11.-03. mp3-file (24:54)
-
Satakunnan Kansa, Esa-Pekka Avela, ``Kännykästä ehkä lintujen tunnistaja'', 16.02.-05. pdf-file
-
Satakunnan Radio, Pertti Kalinainen, ``Satakunan Luontoradio'', 15.03.-05. mp3-file (7:37)
-
Aamulehti, ``Porissa kehitetään linnun äänen perusteella tunnistavaa vekotinta'', 29.05.-05. pdf-file
-
Etelä-Saimaa, ``Linnunlaulun tunnistuskone kelpaisi niin maallikolle kuin biologillekin'', 30.05.-05. pdf-file
-
Satakunnan Radio, Matti Kauvo, ``Luova Tauko'', 09.06.-05. mp3-file (13:26)
|
| Latest results (in Finnish mainly) |
In results, the first column gives an abbreviation derived from the latin name and the
finnish name. The second and third column gives the numbers of individual syllables in
a song and the percentage of correctly identified syllables, respectively. Fourth
column gives the recognition result (this should be the same species as in the first column). The last column gives the second closest guess of the species. Each row represents
an indivisual recording (a bird) in the database.
|
|
|
|
