“The ubiquity of smartphones combined with the power of new machine learning algorithms presents an opportunity to promote positive interactions between humans and birds and thus create new possibilities for avian research,” said lead author Dr. Connor Wood from the K. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology and his colleagues.
“We present the BirdNET app, a free program capable of identifying over 3,000 bird species by sound alone.”
BirdNET uses artificial intelligence to automatically identify the species by sound.
It allows users to record audio on a smartphone and transmit that audio and metadata (date, time, and location) to the BirdNET server, and a bird species identification is provided with a qualitative confidence score.
The raw audio, quantitative confidence score, and metadata are saved on the server for subsequent research usage. All observations are anonymized, and no user data are stored.
“Our guiding design principles were that we needed an accurate algorithm and a simple user interface. Otherwise, users would not return to the app,” explained co-author Dr. Stefan Kahl, also from the K. Lisa Yang Center for Conservation Bioacoustics at the Cornell Lab of Ornithology and his colleagues.
To test whether the app could generate reliable scientific data, the authors selected four test cases in which conventional research had already provided robust answers.
Their results show that BirdNET app data successfully replicated known patterns of song dialects in North American and European songbirds and accurately mapped a bird migration on both continents.
more...
http://www.sci-news.com/biology/birdnet-app-10951.html