Cornell Creates Website To ID Birds From Photos

A new website created by Cornell University in collaboration with the Visipedia research project is going to appeal to birdwatchers in North America, particularly those who might come across a winged fellow of a species unknown to them. The website created by Cornell is going to make it very easy for birdwatchers to find out the species, it will be able to identify the bird’s species only through the photo that’s uploaded by the user to the website.

It’s going to rely upon some sophisticated artificial intelligence magic to identify birds from photos uploaded by users, the only catch is that it must be species that’s found in the United States and Canada.

When a user uploads a picture to the website they will need to draw a box around the bird and click on its tail, beak and eye. They will also be required to mention where and when the picture was taken.

After that information is put in the website will do it’s thing, look through countless pictures taken by other birdwatchers and even the more than 70 million images found on eBird.org to locate the one that matches that of the user.

Users will then be provided with a list of best matches accompanied by photos and recordings of sounds the birds make and the songs they sing.

The website, called Merlin Bird Photo ID, is open to all and free to use.

Adnan Farooqui
Categories: Web
Tags: artificial intelligence