Microsoft’s Project Oxford launched its API in a beta stage yesterday. The project will allow developers to access very advanced API that will allow for facial detection and recognition that can be added directly to developer apps.
Using face detection, the project can even estimate a person’s age.
The API’s main offerings at the moment are: face verification, the ability to find similar faces, image face verification using two images and speech recognition. The project has integrated vision API that will crop images automatically so that a person’s face is always cropped in the center of a photo.
There is a fourth API that also allows for custom language understanding and will be able to be integrated into developer apps.
Microsoft unveiled the project with the how-old.net domain yesterday. The site allows users to upload their photos and uses the project’s API to guess a person’s age. With my own experience, the project was highly accurate with one picture displaying the right age, and the other estimating a year or two older.
The website went viral, and Microsoft linked to the story behind the demo that explains how the demo was produced, and how developers can use Project Oxford for their own applications.