If Intel Labs staff and an Arizona State researcher are successful, you may eventually hold in your hand a TV remote that can know your viewing habits as soon as you grab it, giving you more personalized viewing options and helping block inappropriate TV content from your young children.
The scientists are trying to achieve that goal without needing the user to log in somehow to identify him/herself or without using a face-recognition camera. Instead, the researchers say, “[W]e use the corresponding remote control data, such as a sequence of button presses and accelerometer readings, to train the predictor of the person.” They’ve tested their methodology on a sample set of five households, capturing gestures via accelerometer as well as button pushes and then running the results through algorithms to determine the identity of the remote holder.
The bad news, if you’re getting excited about the prospect of this learning remote, is that the technology does not approach 100-percent accuracy yet. The researchers hope to improve that accuracy by broadening the data set to include the viewing habits of test subjects. As more users log onto the Internet through their HDTVs, their online profiles can also help the remote identify its holder. Eventually, such a remote could help you cut through channel clutter to find exactly what you want to watch — who could root against that?
[Via Engadget]
