Monday, April 24, 2006

Software Is 64% Accurate At Detecting Boredom

Researchers at Massachusetts Institute of Technology are working on a high-tech device with seemingly a multitude of uses in lessening our crushing overload of banality: a boredom detector. A talker, via a wearable camera and software that measures facial expressions and movements, could know whether he has lost touch with a listener (via signals from eyebrows, lips, nose, etc.).

The Emotional Social Intelligence Prosthetic Device has been developed by MIT's Media Lab in order to help people with autism keep track of how they are being received by those around them. The device was created in large part to alleviate the discomfort that autistic people feel when those they are talking to suddenly want to leave. One of the ways that people with autism struggle is by being unable to see such things as people showing boredom or confusion.

The device is a camera, which is small enough to be attached to a pair of glasses, that is attached to a hand-held computer which is running software that recognizes images and the emotions that those images suggest. In a recent test, the software was demonstrated to show whether a person was agreeing, disagreeing, concentrating, or showing disinterest, from only a few seconds of video footage. So if someone you are talking to is acting bored or disinterested, the computer vibrates, letting the camera-wearer know what is going on.

So far, the software is said to be accurate 64 percent of the time, according to a March report in New Scientist.