Thursday 28 June 2018

Bird Identification Challenge

Another challenge, this time on identifying bird calls, with two £250 prizes:

"The task is to design a system that, given a short audio recording, returns a binary decision for the presence/absence of bird sound (bird sound of any kind). The output can be just "0" or "1", but we encourage weighted/probability outputs in the continuous range [0,1] for the purposes of evaluation. For the main assessment we will use the well-known "Area Under the ROC Curve" (AUC) measure of classification performance.
An important goal of this task is generalisation to new conditions. To explore this we provide 3 separate development datasets, and 3 evaluation datasets, each recorded under differing conditions. The datasets will have different balances of positive/negative cases, different bird species, different background sounds, different recording equipment. To solve this task well, you will need an approach which either inherently generalises across conditions (including conditions not seen in the training data), or which can self-adapt to new datasets ("domain adaptation").

Note that for every audio clip, you will be told which dataset it belongs to. This means that adapting to the overall characteristics of each dataset separately is possible. The evaluation will also consider each dataset separately and combine the outcomes, rather than treating them as a single pooled dataset."

Challenge link: http://dcase.community/challenge2018/task-bird-audio-detection

>spectro(2018)

From Jerome Sueur:

Dear all, 

It is my pleasure to announce the contest >spectro(2018), a friendly and informal contest for the best spectrogram generated with R code.

The spectrogram is a 2D/3D key visualisation tool for bioacoustics, ecoacoustics and other sound related disciplines. The spectrogram is not only useful for science, it can also be a nice graphical object with delicate shapes and colours. 
The aim of this contest is to share the beautiful sounds, R codes and spectrograms you may have in your files so that it can help others to produce nice graphics and figures.

But overall, the idea is to join science, fun, and maybe the arts!

Here are some information about the organization of the contest:

. deadline : 15  September 2018
. application: a single submission per candidate; send wav sound + R code + png image to jerome.sueur@mnhn.fr with the email title ">spectro(2018)"
. sound: .wav file / no limits, any recording (animal, habitat, voice, music, etc) or synthetic sound built with R
. code: .r file / only within R language, any package but no call to external software, the code should be fully repeatable
. image: static .png file / no limits in the choice of colour and size but should be readable on a web page
. rights: you accept that your material appears on seewave webpage 
. international voting committee: Fanny Rybak (France), Nadia Pieretti (Italy), Susan Fuller (Australia), Stefanie LaZerte (Canada), Tess Gridley (South Africa)
. publication of the results: October-November 2018
. prize: 1 printed sample of the book Sound analysis and synthesis with R for the winner, 1 electronic version of the same book for the second and the third, kindly offered by Springer, Berlin

Have fun and good luck!

Jerome

ShareThis

Copyright