Feeling good? Ready for the weekend? Pumped about life? Microsoft can change that in a hurry: Just upload your photo to the tech giant’s How-Old.net, so its algorithms can attempt to guess how old you are. There’s nothing quite like a machine-learning platform adding 25 years to one’s actual age.
How-Old.net relies on a machine-learning API called Face API, found in the Azure Machine Learning Gallery. As the name implies, the API extracts information about faces in photographs—including the latitude and longitude of wherever the picture was uploaded.
While Microsoft claims the website doesn’t save the photos, the metadata from each upload is collected via Azure Event Hubs and crunched using Azure Stream Analytics, a low-latency processing platform. Microsoft is showing off the technology in the context of guessing people’s ages, but they could do a lot more with the data, especially in aggregate; it wouldn’t be shocking, for example, if the company’s developers issued a blog posting in the near future that details the average age of a How-Old.net user, for example.
But the technology isn’t perfect: Based on a photo, How-Old.net thinks that Teddy Roosevelt—who died at the age of 60—is a 63-year-old woman.
- Amazon Launches Machine-Learning Platform
- At Build, Microsoft Pushing Windows 10 Hard
- How We Data-Mine Related Tech Skills