BirdNET is a free app from the Cornell Lab of Ornithology that identifies birds from their songs and calls using a deep convolutional neural network. A June 28, 2022 paper in PLOS Biology by Connor Wood, Stefan Kahl, and colleagues at the lab’s K. Lisa Yang Center for Conservation Bioacoustics documented how the tool, which could identify more than 3,000 species at the time, lowered the barrier to bird research by letting anyone contribute sound recordings rather than requiring expert visual identification.
The reach was substantial. In 2020 alone, over 1.1 million participants generated 31 million submissions through BirdNET, of which 5.8 million observations were retained after quality filtering - a scale that dwarfed the participation of comparable expert-oriented platforms. The paper validated the resulting data against four ecological case studies covering song dialects and migration patterns, showing that crowd-sourced AI identifications could support real scientific conclusions.
Why business readers should care: BirdNET is a model for AI as a force multiplier for science. By turning a smartphone microphone into an expert classifier, it converted millions of casual users into a distributed sensor network for biodiversity monitoring, illustrating how a well-targeted classifier can generate research-grade data at a scale no professional team could match.