Merlin App to Feed Bird IDs Into Cornell's eBird Database
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Cornell Lab of Ornithology's Merlin app will let users automatically upload real-time bird sound identifications to the eBird database of 2bn+ observation records, linking ML-based detections to citizen-science monitoring and conservation tools.
- Merlin, which uses ML trained on spectrogram patterns, currently identifies 2,066 bird species and has been downloaded 40m+ times across 240 countries, with Britain ranking second in usage (nearly 2m users in May alone).
- The European Bird Census Council recommends against using Merlin in official breeding bird surveys over misidentification concerns and has set up a monitoring group to coordinate acoustic bird monitoring across Europe.
- RSPB's Richard Gregory warned the app still makes errors — it once identified his dachshund as a mallard — flagging a conservation risk if incorrect species data feeds into eBird.
- UK bird populations have fallen by more than 70 million over the last 50 years per the British Trust for Ornithology, underscoring the conservation stakes behind scaling acoustic citizen-science data.
Why it matters: With UK bird populations down 70 million over 50 years per the BTO, funneling Merlin's 40m+ users' recordings into eBird will scale citizen-science data feeding conservation tools — but the EBCC's exclusion of Merlin from official surveys and RSPB's Gregory flagging misidentifications (his dachshund was tagged as a mallard) show the data's usefulness depends on accuracy improvements.
