Using sound, bird ID app opens a door for citizen scientists

Biologists at the Cornell Lab of Ornithology knew their BirdNET app was popular. Since they rolled it out, over 2.2 million people have used it to identify birds by their songs and contribute their findings. But the app is also accurate enough to provide reliable scientific data. In a study published in PLOS Biology, researchers in the K. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology took four test cases and compared conventional results with BirdNET. The test found that the app was able to accurately replicate bird song dialects across North America and Europe and accurately mapped bird migration on both continents. The app uses machine learning to identify over 3,000 birds by sound. Developers say it can make citizen science more accessible because it doesn’t require bird identification skills to participate. They hope the app can play a role in a long-term global research effort to identify birds and other wildlife.

Read the study here.

Header Image: The BirdNET app identifies birds by sound. Credit: Stefan Kahl/Cornell Lab of Ornithology