Is AI for the birds? The beauty of backyard birdsong data (WHY2025)

Birdsong is all around us, but is there a deeper meaning? Can computer analysis help us decipher these hidden patterns in our own backyards? This talk charts my journey exploring the world of open source projects for bird (and bat!) audio identification, some of the systems I've operated, and the data art pieces I've created from the results. From the science of birdsong, to machine learning models and data visualization, there's something for every hacker to be inspired by. - How birds sing, spectrograms and analysis - Neural networks for identifying animal sounds - Feature pre-processing for sound ID - Open source projects you can build yourself - Data art and visualizations - The psychology of shifting baselines and why the data matters Licensed to the public under https://creativecommons.org/licenses/by/4.0/ about this event: https://program.why2025.org/why2025/talk/L79ASB/