Noise Robust Bird Song Classification, Recognition, and Detection

Noise Robust Bird Song Classification, Recognition, and Detection

-- a summary of Wei Chu's work on bird song processing and analysis

Motivation

Database

Bird songs are important in the communication between birds of specific species. A bird can listen to other birds and classify them as conspecific or heterospecific, neighbor or stranger, mate or non-mate, kin or non-kin [1]. It can also sing to other birds for mate attraction, danger alert, or territory defense [2]. Behavioral and ecological studies could benefit from automatically detecting and identifying species from acoustic recordings.

Database

RMBL-Robin database
A 78 minutes Robin song database collected by using a close-field song meter (www.wildlifeacoustics.com) at the Rocky Mountain Biological Laboratory near Crested Butte, Colorado in the summer of 2009 [3]. The recorded Robin songs are naturally corrupted by different kinds of background noises, such as wind, water and other vocal bird species. Non-target songs may overlap with target songs. Each song usually consists of 2-10 syllables. The timing boundaries and noise conditions of the syllables and songs, and human inferred syllable patterns are annotated.
The database is used for bird song detection. It can be downloaded from here.

To reference the RMBL-Robin database, please use the following:
W. Chu, D.T. Blumstein, “Noise robust bird song detection using syllable pattern-based hidden Markov models,” ICASSP 2011, pp. 345-348.
Please reference RMBL-Robin database when using it.

Antbird database
Researchers from the Ecology and Evolutionary Biology department at UCLA recorded 127 minutes long 3366 calls from 5 species of Antbirds (Barred Antshrike, Dusky Antbird, Great Antshrike, Mexican Antthrush, Dot-winged Antwren) in a Mexican rainforest [4]. Different kinds of background noises are observed in the recordings, such as other bird chirps and insect sounds. The calls are 0.5 - 5.0 seconds long.
The database is used for bird species classfication. If interested in using it, please contact Prof. Charles Taylor.

Publications

• W. Chu and A. Alwan, “fbEM: a filter bank EM algorithm for the joint optimization of features and acoustic model parameters in bird call classification,” Interspeech 2012, pp. 1993-1996. [poster]

• W. Chu, D.T. Blumstein, “Noise robust bird song detection using syllable pattern-based hidden Markov models,” ICASSP 2011, pp. 345-348. [poster] [database]

• W. Chu, A. Alwan, “A correlation-maximization denoising filter used as an enhancement frontend for noise robust bird call classification,” InterSpeech 2009, pp. 2831-2834. [slides]

Acknowledgement

Supported in part by NSF.

References

[1] P. Marler, “A comparative approach to vocal learning: song development in white-crowned sparrows,” J Comp Physiol Psychol, vol. 71, pp. 1–25, 1970.
[2] C. K. Catchpole and P. J. B. Slater, Bird Song: Biological Themes and Variations, Cambridge University Press, New York, 1995.
[3] W. Chu and D.T. Blumstein, “Noise robust bird song detection using syllable pattern-based hidden Markov models,” ICASSP 2011, pp. 345-348. [poster]
[4] V. Trifa, A. Kirschel, and C. E. Taylor, “Automated species recognition of antbirds in aMexican rainforest using hidden Markov models,” The Journal of the Acoustical Society of America, vol. 123, no. 4, pp. 2424–2431, 2008.

Contacts

If you have any questions or suggestions, feel free to contact Wei Chu (weichu@ucla.edu).

last updated: May 26, 2011.