Shortly after Microsoft released How Old Do I Look, an gender-guessing, age-telling app that while widely popular was often wildly wrong (it was hit and miss with our own editorial masthead), Redmond has now released Twins or Not.

Built on the same principles of How Old Do I Look Twins or Not uses machine learning APIs to make an assertion about the face (this times faces) in the photos. Either take two photos from the web, or upload from your own computer, and the app will tell you how closely you are related from what it can determine about your facial features.

Matt Velosso, who created the app in a matter of a few hours, described in a blog post how it was done, somewhat focusing on how a “viral” site can be made with ease in just a short time (even if it’s not that accurate, it’s at least amusing): “1-I signed up for Azure, Microsoft’s cloud platform. 2-Obtained an API key for the Face API, which is a subset of our Machine Learning APIs focused on facial recognition. 3-Downloaded the Face API SDK in .Net which already has a sample that does almost exactly what I needed. 4-Adapted the code into an ASP.Net MVC website. 5-Published the site to Azure.”

True enough the site had many thousands of users just a few hours after it went up, although mobile devices proved to be problematic. After giving Twins or Not a shot at guessing how close two pictures of myself were it wasn’t bad at all, given that the photos were taken six years apart. The result: “Twin score: 100% OMG, clones!!!” It wasn’t always successful though as scores of Twitter users testifed, and while some users wrote that such apps were proof that Microsoft was “cool again”, others were not so kind (“This thing continues to be crock full of burns”).

Photo credit: MarciPlan, Twitter