New site built with Azure services helps you find your twin

New site built with Azure services helps you find your twin

Last updated: October 25, 2017 

Have you always felt a void in your life? Were you separated at birth? A new site helps you find your twin — as long as you already have their photo. Fresh off the successful heals of how-old.net, the site was introduced to use technology to help you find your twin. Sort of.

When two photos are uploaded to TwinsOrNot.net, they are compared, and assigned a Twin Score. The Twin Score is essentially a resemblance percentage. When the Twin Score percentage is low, there is little to no resemblance between those pictured in the two photographs. When the Twin Score percentage is high, the site says the two are definitely related, or quite possibly clones. The uploaded photos are not retained.

You may recall the age-guessing website was demoed during Microsoft’s BUILD 2015 developer conference in April to highlight the capabilities of Microsoft’s Azure services. It quickly went viral. That site guesses the age and gender of the people in an uploaded photograph. Like the age-guessing site, the twin-finding site, TwinsOrNot.net, was also built to highlight the capabilities of Azure services using an API from Microsoft’s Machine Learning APIs.

The site was created by Mat Velloso, a Senior Software Development Engineer at the Technical Evangelism Development group at Microsoft. According to Mat Velloso’s blog, the site was built in about 4 hours in a hotel room using the Face API in Project Oxford:

This site takes two photos and will assess how similar the people in these photos are, giving you a score from 0 to a 100.  In order to do this, I used the Face API in Project Oxford, a platform of intelligent services. This is just one of the many other amazing features our Machine Learning APIs and platform can offer.

Thank you for visiting Tech Help Knowledgebase to learn how to find your twin.