The hills are alive with the sound of birds. Awaking to the dawn chorus is an immersive and wonderous experience, but it's difficult to capture via a passive monitor with the fidelity needed for research-grade classification. In recent years, AI classification models like BirdNET have seen incredible advancements, and while consumer devices now leverage this technology, little has been done to improve the capture quality—it's typically still limited to a single microphone.

Simultaneously, beamforming and source-separation have enabled high-quality capture for human voice, primarily focused on conference room environments. Human voice occupies a significantly lower frequency range than most passerines, however, meaning the mic arrays currently on the market are not sufficient for separating the higher frequencies of concurrent birdsong.

This is the problem we are solving. We are leveraging and connecting technological advancements across acoustics and machine learning with a dedicated focus on avian research, enabling a new level of accuracy and fidelity in passive field monitoring of bird populations.