WhaleID is the first AI system built for underwater whale identification. It uses 7 neural networks to analyze full-body underwater humpback whale photos from dive and snorkel encounters — head tubercle patterns, pectoral fin scarring, body markings, dorsal fin shape, and fluke trailing edges.
Unlike traditional fluke-only matching, WhaleID's anatomical approach works from any viewing angle using our detection system, cross-view network, identity network, and feature network extraction.
Most whale-ID tools match from a single feature — typically the tail fluke. WhaleID analyzes six regions — head, fluke, pectoral fin, body, dorsal fin, and caudal peduncle — using seven neural networks, so a match is possible from any viewing angle rather than a fluke-up shot alone.
A free research platform using 7 neural networks to analyze humpback whale photos from any body part, including head tubercle patterns, pectoral fin scarring, body markings, dorsal fin shape, and fluke trailing edges.
Most whale-ID tools match from the tail fluke alone. WhaleID analyzes head, fluke, pectoral, body, dorsal, and peduncle regions from any angle using proprietary detection and custom biology-informed extractors.
Yes, completely free for individuals, whale watching companies, and researchers.
Upload whale photos at whaleid.org/submit. No account required.
Built by Alex Roessner. About the Science | Explore Gallery | Contribute a Photo