Cornell Opens Archive of Bird Calls Going Back to 1929

Awesome. I've checked out the Macaulay library a few times in the past looking for samples for electronic music. It's a great resource, and has features like rating by audio quality. Check out the unearthly vocalizations of various Oropendolas:

http://macaulaylibrary.org/search?&asset_format_id=1000&coll...

At around 0:25 in this recording is an example of what I mean, but each member of the genus has its own distinct calls/songs:

http://macaulaylibrary.org/audio/12830

First thing I thought was: let's download that and make a game out of it; maybe use location data to start of easy with birds from the region, but I'm afraid it can't be used for commercial purposes (all mobile apps are classified as such by their documentation) without acquiring a license (which is understandable, but regrettable at the same time).;

Would be great if there were a tensorflow project that would learn the calls and then could be used to identify birds in the wild :)

I couldn't wait and a few years ago scraped their site: http://blog.databigbang.com/the-call-of-the-web-scraper/

if you like bird calls, take a second to visit xeno-canto.org, superb amateur birder site that collects audio of bird calls, via (i believe) an cc artist license and is totally free. Also, run by Dutch mathematicians!

Reminds me of this algorithm competition to classify bird species from recordings: https://www.kaggle.com/c/mlsp-2013-birds

Dead relevant comment in this thread: https://news.ycombinator.com/item?id=12156328