Monday, April 16, 2018

Cuckoo Clocked: Can an app make you a better birder?

Cuckoo Clocked: Can an app make you a better birder?

Cornell Lab’s Merlin app is coming to India. They’re crowdsourcing images from local birdwatchers, and everyone’s aflutter.

The Merlin Bird ID app has changed birding in North America in the four years since its launch, helping novices identify species in seconds, from a single photo.
The Merlin Bird ID app has changed birding in North America in the four years since its launch, helping novices identify species in seconds, from a single photo.(Image Courtesy Cornell Lab)
In theory, it sounds perfect. You spot a bird in the wild (or on the windowsill of your concrete jungle), take a quick shot with your smartphone, and an app identifies the species in seconds with 90% accuracy.
For those who’ve used it, four-year-old Merlin Bird ID has revolutionised birdwatching. But the app, developed by the Cornell Lab of Ornithology and Caltech, using crowdsourced photographs, is perhaps better known for how academia, machine learning and crowdsourced data can collaborate towards conservation. In North America, where it was launched, it’s allowed ordinary folks to identify birds without bulky field guides or long-drawn-out Google searches. Among established birdwatching groups, it’s cut down endless debates on which bird was spotted. Naturalists the world over have been using the data to understand migratory patterns, habitat changes and other avian issues.
In January, the Cornell Lab announced plans to extend the app for birds in India – a decision that is as exciting as it is daunting. India ranks among the world’s 12 megadiversity nations, with 1,266 or 13% of the world’s bird species. But for Merlin to identify Indian birds, it needs more than 500 reference photos for every species. Local birders have been urged to contribute their pictures to build the database.
Some birds like the bright blue verditer flycatcher are easy enough – contributions have already crossed 750. Others like the blue-yellow Banasura Laughingthrush are so rare, there are only seven pictures uploaded.
India’s birdwatching community is divided over the use of tech-driven tools like bird-identification apps. Some say it will popularise the hobby. Others fear it will become a distraction, reducing the activity to almost a game. (Pramod Thakur / HT File Photo)
SPREADING WINGS
Mohit Aggarwal, 31, a bank executive who has lived in several cities and is now based in Mumbai, has contributed over 100 images, most of rare species. “I’ve focused on filling the gaps left by other contributors,” says the birder.
He hopes the app will soothe the ruffled feathers of India’s birding community. “A unique sighting usually causes controversy,” he says. Older birdwatchers, who’ve been bird-spotting before digital cameras, tend to be dismissive, even disbelieving, of younger enthusiasts’ pictures, he says. “Photos offer proof that a certain bird has been in an unlikely region. When a picture is up for identification, you’re typically up against someone’s ego. Machine learning may be able to answer without bias.”
For Albin Jacob, 36, a software engineer from Bengaluru, contributing more than 3,000 pictures was a breeze. He’s photographed more than 800 species across India and is a reviewer for the India portal of Ebird, Cornell Lab’s massive crowdsourced database of bird observations.
“I’m excited that the app will be available for India,” he says.
Source: Hindustan Times dated April 15, 2018

Featured Posts

Top Searches from “IEEE Xplore Digital Library" - 13th September 2024

  The Learning and Information Resource Centre is pleased to inform you about the Top Searches from  "  IEEE   Xplore   Digital Library...