This article was published on February 12, 2016

Microsoft’s new app can identify dog breeds from photos


Microsoft’s new app can identify dog breeds from photos

Over the past few months, Microsoft has released a slew of apps that use its Project Oxford AI tech to guess from photos how old you are, what emotions you’re experiencing and how long your moustache is.

Today, the company has launched Fetch!, a Web and iOS app that can identify dog breeds from pictures. And it’s pretty darn accurate.

I chucked the desktop version of Fetch! a close-up of my Bassett Hound’s face and it correctly guessed her breed. It also got a few others right, including a bull terrier and a Cavalier King Charles Spaniel.

The app isn’t perfect: it was wrong a few times, but only by a little. It guessed very similar breeds when I uploaded images of a Bullmastiff and a Rajapalyam, which is a racing dog native to India.

The <3 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol' founder Boris, and some questionable AI art. It's free, every week, in your inbox. Sign up now!

Microsoft says its iOS app offers even more features, like a percentage meter showing how close your uploaded picture matches the breed in its database, along with info on several breeds.

Fetch's iOS app has info on a range of breeds and can also tell you what breed you resemble
Fetch’s iOS app has info on a range of breeds and can also tell you what breed you resemble

Oh, and it can also look at photos of humans and tell you what breed they resemble.

Fetch! is a product of Microsoft Garage, the company’s program to bring employees’ experimental projects to life. It has previously released an alarm clock app that makes waking up fun, a tool for taking and editing selfies and another for planning meetups with friends.

Fetch! New Microsoft Garage app uses artificial intelligence to name that breed [Microsoft News Center]

Get the TNW newsletter

Get the most important tech news in your inbox each week.

Also tagged with


Published
Back to top