Prisons Testing Computers Capable of Detecting, Defusing Riots
As the New York Times explains, one such system uses artificial intelligence software to instantly analyze images fed from security cameras. The software is capable of recognizing the faces, gestures and movements of inmates, and can automatically pick up on signs of danger (e.g., if one group of prisoners begins moving toward another). Once the system senses trouble, it will automatically send a text-message alert to a correctional facilities officer, who can move in and cool tempers.
The system is currently being tested at correctional facilities across the nation, and represents the latest frontier in the fast-growing field of computer vision technology. DARPA, for example, has already launched a program called 'Mind's Eye,' which involves new machines trained to analyze and communicate everything they see. Meanwhile, a hospital in Cooperstown, NY has begun experimenting with similarly observant cameras, which can monitor a nurses' movements, and remind them to wash their hands, if they haven't already. And, of course, the Kinect add-on to Microsoft's Xbox 360 has already brought computerized vision to the living room.
Under certain circumstances, the benefits of this new technology seem self-evident. Prison guards can devote their time to other duties, patients can rest assured that their doctor has followed protocol, and, in extreme cases, police could even use enhanced surveillance cameras to track a criminal's whereabouts. Privacy advocates, however, have expressed some major concerns--and with good reason.
It's one thing to see security cameras perched atop airports or shopping malls. It's quite another to know that any camera could be collecting far more information about you than you'd ever imagined. Can it tell what products you like to consume? Can it see that you hate your boss? Or, more creepily, can it even identify you by name, address, or occupation? It's questions like these that have forced many companies to tread carefully into the world of computer vision. But we likely won't know where the technology leads until it more fully matures. "With every technology, there is a dark side," said Hany Farid, a computer scientist at Dartmouth. "Sometimes you can predict it, but often you can't."