Merlin Bird ID
Designed for anyone interested in identifying birds, Merlin is a machine learning-powered bird ID tool created by the Cornell Lab of Ornithology. As a global bird guide, it can help you identify birds anywhere in the world.
Machine Learning Magic
Merlin uses computer vision and machine learning technology to identify birds in photos and sounds. Merlin learns to recognize bird species based on sets of millions of photos and sounds submitted to eBird by birders around the world, archived in the Macaulay Library at the Cornell Lab.
Photo ID was developed by Dr. Grant Van Horn while a member of Dr. Pietro Perona’s computational vision lab at Caltech, and Dr. Serge Belongie’s computer vision group at Cornell Tech, collaborators on the Visipedia project. Merlin Sound ID was developed by a team at the Macaulay Library, led by Dr. Grant Van Horn.
Merlin is powered by birders like you! Merlin taps into over a billion records in the eBird citizen-science database to find out which birds are likely to be found in your vicinity on a given date. The 87,000+ photos and sounds featured in Merlin are from thousands of photographers and sound recordists who submit their media to the Macaulay Library through eBird. Merlin delivers accurate results thanks to the experienced birders who curate and annotate the photos and sounds used to train Photo ID and Sound ID.
Merlin is made possible by support from the Macaulay Family Foundation, the National Science Foundation (grant number DRL-1010818), the Faucett Catalyst Fund, Swarovski Optik, Pennington® Wild Bird Food, and friends and members of the Cornell Lab of Ornithology. Merlin Photo ID, powered by Visipedia, was made possible by support from Google and the Jacobs Technion-Cornell Institute.
We thank these supporters and the entire community of birders around the world that helped build Merlin.