The goal of BirdNET is to build an automated identification system for birds based on machine learning. It can identify up to 3000 common species of birds that hail in North America and Europe. For each species, the system learns to recognize a unique "sound signature" based on audio recordings from various species. The resulting identification system can identify all birds at once, regardless of species.

You can learn about the different species at the same time. Record your own sound clips and submit them to the system for analysis.