ISCA Archive Interspeech 2015
ISCA Archive Interspeech 2015

Quantifying difference in vocalizations of bird populations

Colm O'Reilly, Nicola M. Marples, David J. Kelly, Naomi Harte

Populations of a bird species can evolve over time to become a new species. While plumage patterns and other morphological information can remain constant, the vocalizations of a given population may have diversified enough to warrant reclassification. Thus ornithologists are interested in how to measure call and song similarity in birds in a systematic, repeatable manner. Given the success of speech processing methods applied to bird species classification, this paper presents work on developing a measure of bird call similarity. The method is inspired by human speech dialect separation measurement using a representation of the pitch contour micro-structure. The measure is applied to bird populations with calls that are considered very similar, very different and between these two extremes. Initial results are very promising, with the behavior of the metric consistent with accepted levels of similarity for the populations tested to date.

doi: 10.21437/Interspeech.2015-677

Cite as: O'Reilly, C., Marples, N.M., Kelly, D.J., Harte, N. (2015) Quantifying difference in vocalizations of bird populations. Proc. Interspeech 2015, 3417-3421, doi: 10.21437/Interspeech.2015-677

  author={Colm O'Reilly and Nicola M. Marples and David J. Kelly and Naomi Harte},
  title={{Quantifying difference in vocalizations of bird populations}},
  booktitle={Proc. Interspeech 2015},