Access to software that performs complex operations has been, and can still be, costly. Paying for expensive licenses can mean an alternative solution has to be sought or worse still, a feature won't be developed at all. Thanks to virtualization and a many other cloud features, we've seen these complex operations offered as services at a fraction of the cost. One such complex operation is that of facial recognition. This software hasn't seemed accessible at all with only applications such as Google Picasa, Apple's iPhoto, Facebook and more recently XBox Kinect being the most obvious examples where it has been used. These are big companies with big money to spend. This now changes as Face.com is offering a free to use face recognition API for photos.
The API has already been used by 10,000 developers during alpha testing, where there was a restriction of scanning 200 photos per-hour. In Beta the API limit will be upped significantly to a 5,000 photos per-hour limit. The service also boasts features such as grouping of similar faces, bulk-tagging, integration with Facebook (recognize Facebook friends in photos on Facebook Connect apps) and Twitter integration (search and tag Twitter faces across photo services). Face.com have developed Facebook applications including Phototagger and Photofinder (in Alpha) and the technology is in use by EyeBuyDirect, AXE (Lynx in the UK) and Ralph Lauren.
While the Face.com website and documentation consistently mention "photos," it would be interesting to consider things such as analyzing video capture stills for almost real-time facial recognition. How about giving your local café the permission to auto-check you in on foursquare whenever their security camera sees you enter?
We would be interested to hear if you have used the Face.com alpha API, are trying out the Beta API or have any thoughts about uses for this service. Please let us know your findings either in the comments below or on the Face.com API profile.