Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Multi-Label Classifier Chains for Bird Sound

May 29, 2013
Forrest Briggs, Xiaoli Z. Fern, Jed Irvine

Share this with someone who'll enjoy it:

Bird sound data collected with unattended microphones for automatic surveys, or mobile devices for citizen science, typically contain multiple simultaneously vocalizing birds of different species. However, few works have considered the multi-label structure in birdsong. We propose to use an ensemble of classifier chains combined with a histogram-of-segments representation for multi-label classification of birdsong. The proposed method is compared with binary relevance and three multi-instance multi-label learning (MIML) algorithms from prior work (which focus more on structure in the sound, and less on structure in the label sets). Experiments are conducted on two real-world birdsong datasets, and show that the proposed method usually outperforms binary relevance (using the same features and base-classifier), and is better in some cases and worse in others compared to the MIML algorithms.

* 6 pages, 1 figure, submission to ICML 2013 workshop on bioacoustics. Note: this is a minor revision- the blind submission format has been replaced with one that shows author names, and a few corrections have been made 

   Access Paper Source

Share this with someone who'll enjoy it: