14 DAY TRIAL // If you have a hard time telling how old a person you recently met actually is, Microsoft comes to the rescue with a service that can guess the age of any human being. More or less.
Designed as an internal project whose development took only one day, how-old.net is a service that uses machine learning to tell how old a person shown in a photo is, analyzing a number of factors that could help determine gender and age.
The website allows users to upload any photo and let the API try to predict the age of any face recognized in the picture. Obviously, the result won’t be 100 percent accurate, but in most of the cases, it gets pretty close with a margin of 4 or 5 years.
Redmond says that it was all designed to be just an internal project, so it expected approximately 50 people to try it out, but it seems that how-old.net has become an Internet sensation in the last 48 hours.
“We were shocked. Within a few hours, over 35,000 users had hit the page from all over the world (about 29k of them from Turkey, as it turned out – apparently there were a bunch of tweets from Turkey mentioning this page). What a great example of people having fun thanks to the power of ML!” Microsoft says.
How to use it
How-old.net is really easy to use. All you have to do is open the website, upload your own photo or search the web for one and the system automatically detects any face that might exist in the picture and tries to determine the age of the person.
There’s also a search box if you simply want to search for faces online with Bing, so if you do want to have some fun guessing how old famous people are, this website could really be a great time waster. But remember that results aren’t 100 percent accurate.