Hi, all!

We're launching the new BirdCLEF+ competition presently - the 'plus' is because 
we've got multi-taxa data this year, including amphibians, insects, mammals, 
and, of course, birds. This has been a major annotation undertaking, led by 
Juan Sebastián Cañas of the University College London and Humboldt Institute in 
Colombia.

https://www.kaggle.com/competitions/birdclef-2025

I'm writing to you because you might have students who could be interested in 
participating! Here's a bunch of great things about the competition, especially 
for students:

  *   We have $50k in total prize money!
  *   People can submit 'working notes' for the CLEF competition, detailing 
their approaches. There's $5k in prize money for the best working note, and 
it's a great chance to get some practice with academic writing. The working 
notes are published in the proceedings of the CLEF conference, and the best 
ones do get more widely 
cited<https://scholar.google.com/scholar?hl=en&as_sdt=0%2C5&q=mario+lasseck+birdclef&btnG=>,
 as well.
  *   Often, the best working note is not one of the top-scoring entries! We 
like to reward well-documented, creative approaches.
  *   Kaggle is a great learning platform - students can get hands-on 
experience attacking a problem as part of a community which I find to be 
surprisingly helpful and kind.

And a few of my favorite features of the competition itself:

  *   We include a sizeable tranche of unlabeled in-domain data: this is a 
great test bed for domain-adaptation methods.
  *   Each year, we work with a different organization to develop a new 'fully 
annotated' dataset, and we publish the test dataset after the competition is 
complete. This has steadily increased the number of regions where we have some 
high-quality groundtruth for bird species identification. In fact, the BirdSet 
benchmark<https://arxiv.org/pdf/2403.10380> is mostly an omnibus of past 
BirdClef test data sets.
  *   We use a class-averaged ROC-AUC metric, which focuses on core 
classification quality, rather than thresholding methods.
  *   And, as mentioned up top, this is an exciting year because we're moving 
to a multi-taxa task.

So, if you've got students who might be interested in taking some time for a 
bioacoustics competition, do send them along!

Cheers,

Holger

--

At the Yang Center, we work flexibly, and while it suits me to email now, I 
don’t anticipate a response outside your regular work hours. Thanks!

--

Dr. Holger Klinck

John W. Fitzpatrick Director

K. Lisa Yang Center for Conservation Bioacoustics

Cornell Lab of Ornithology, Cornell University



Faculty Fellow

Atkinson Center for a Sustainable Future, Cornell University



Courtesy Professor

Marine Mammal Institute

Department of Fisheries, Wildlife, and Conservation Sciences, Oregon State 
University



Mailing address:

159 Sapsucker Woods Road

Ithaca, NY 14850, USA



Tel:    +1.607.254.6250

Fax:    +1.607.254.2460

Email:  [email protected]<mailto:[email protected]>

Web:    https://bioacoustics.cornell.edu<https://bioacoustics.cornell.edu/>



How I say my name<https://www.name-coach.com/holger-klinck>

Pronouns in use: He/Him/His


Reply via email to