The problem with a multi-user tabletop computer is that it's difficult for the software to keep track of who's using it. Unless you strap a Kinect sensor to the underside that lets it make note of their footwear choices.
Developed by the Human-Computer Interaction group at the Hasso-Plattner-Institut in Germany, Bootstrapper uses a Microsoft Kinect sensor to isolate and photograph each user's shoes. At the same time the software keeps track of where hands have touched the top of the table, making educated guesses as to how many users there are, where they're standing, and what kicks they're wearing.
All of this information is then used to not only keep track of an individual user's activities, but to also automatically remember who they are when they return to the table. And so far with a database of 18 different users it's been successful in identifying them about 89 percent of the time. It's of course not a perfect solution—it can probably be fooled in a corporate setting where everyone's wearing black dress shoes—but as part of a larger approach to autonomously recognising users it's certainly a novel approach. [Hasso-Plattner-Institut via MIT Technology Review]