I hate it when I can’t find my TV remote. Once I even managed to leave it in the refrigerator (don’t ask) and couldn’t find it for three days.
But the days of TV remote losing may soon be over, if a UC San Diego computer science grad student named Jacob Whitehill has his way. He’s invented a system that allows him to control video playback … with his face.
Whitehall studied the facial expressions that students make while watching a recorded video lectures. Some expressions, like nodding, suggest comprehension while others, like not blinking as much (strange but evidently true) indicate confusion. The idea is to more closely study how facial expressions correlate to comprehension and confusion, then use face-recognition technology to translate those expressions into data used to control video playback.
So if this all pans out, a student could be watching a video lecture on quantum physics, say, and based on his or her facial expressions, the video go play at regular speed, or back up to repeat stuff that the student finds confusing. Check out a video of what this might look like here.
One problem is that, according to Whitehall, facial expressions among students vary widely. One guy’s puzzled expression is another’s look of boredom, while yet another viewer’s slack-jawed visage is a sign of intense interest and complete absorption. So unless there really are universal facial signs of comprehension and confusion, it could be difficult for this technology to work to help students learn.
But it would seem pretty simple to rig up something that would actually allow you to use your face as a remote control. There could be a little face-recognition box on top of the TV that you could program to respond to facial commands. A nod upwards would mean scroll the channel up, and nod down would mean scroll down. A head jerk to the left could be rewind, a head jerk to the right fast forward. And while we’re at it, why not incorporate voice-recognition software, too? How cool would it be to be able to simply speak a channel or command and make it happen. I don’t know much about face and voice recognition technology, but doesn’t this seem feasible? I think I’m on to something here … (Actually, there’s already something like this–a voice activated TV remote control. Check it out here.)
the reality of the motion detection interface might be difficult as other gestural activities also happen when watching tv .
not all people watch in isolation and also communicate with others while watching .. so the pattern of the movements would appear to chaotic.otherwsie they would have to sit transfixed.
ie watching tv eating popcorn turn your head and dam the program changed.
just a though ..sound activation appears more feasible.
In order for facial expressions to be useful, they would have to be very clear that the intent of the user was being carried out and it would have to be an unambiguous type of indication. Perhaps it would need training like the voice recognition systems used to need.