Posted 05.19.2011 at 4:14 pm
5 Comments
Kinect Recognizes Sign Language
Herbert Wassner
A group of French Kinect hackers/developers (when you have the blessing of the company, is it really hacking?) proved that the Kinect is capable of understanding much more than your barked commands to pause Netflix: It can also understand sign language.
So far, the hack is pretty limited--we're kind of doubtful that the Kinect's relatively low-resolution cameras can pick up the high level of detail that regular-speed American Sign Language entails, but at least we've now got evidence that it can understand some of the more broad gestures. The hack in its current form can only recognize "hello" and "sorry," both of which are made up of exaggerated, highly visible movements, but it seems to do so flawlessly.
Plus, the developers (warning: French) expect to expand the Kinect's vocabulary very rapidly; the framework is all in place, so all they have to do is input the gestures and their corresponding meanings. But it also means that this particular system can be used for all kinds of other things: Basically, it's a simple translator of motions into meanings, so theoretically it could be used for all kinds of other commands. In the meantime, it could be a useful tool for teaching sign language, forcing the student to master each command enough for the Kinect to understand it.
[via Kotaku]
Want to read more articles like this, plus tips and tricks, home hacks, DIY projects, and more? Subscribe to Popular Science today, for less than $1 per issue!
5 Comments
To comment, please Login.
Popular Tags
Regular Features
|
|
|
|
|
|
June 2013: American Energy Independence
Five amazing, clean technologies that will set us free, in this month's energy-focused issue. Also: how to build a better bomb detector, the robotic toys that are raising your children, a human catapult, the world's smallest arcade, and much more.
Popular on Popsci
Most Commented
Most Emailed
Online Content Director: Suzanne LaBarre | Email
Senior Editor: Paul Adams | Email
Associate Editor: Dan Nosowitz | Email
Assistant Editor: Colin Lecher | Email
Assistant Editor: Rose Pastore | Email
Contributing Writers:
Kelsey D. Atherton | Email
Francie Diep | Email
Shaunacy Ferro | Email
Today on PopSci.com
Copyright © 2012 Popular Science
A Bonnier Corporation Company. All rights reserved. Reproduction in whole or in part without permission is prohibited.
I like to make a interesting observtion. Nobody as yet made any comments artical. Could it be a reflection, nobody is impress or they do not care, hmm?!
lol bubba i think they dont care anymore. seeing how the kinect can do just about anything. and honestly sign language isnt that impressive.
There is so much more to ASL than hand gestures: facial expressions, body movements, eye movements and pre-establishing roles all play crucial parts in signed communication. True ASL, if translated directly to english, would make no sense whatsoever. This "hack", while neat, will be very primitive at best and would only be able to teach extremely basic ASL signs, in no way paving a strong foundation.
bigburb is correct. ASL requires so many different visual cues, that this will only work for a very few signs. Many signs use the same hand shape, and potentially even the same movement, but a different facial expression. The best example of this is "here" vs "what". The only difference is the facial expression.
That being said, this is still pretty cool. Because when someone makes a more advanced Kinect system (read: better cameras) it will be able to understand ASL much better.
It has been shown that Kinect can recognize facial expressions (ish, pre established roles can be understood easy enough, eye movement sounds like the kind of thing to cause the most problems here what I don't understand is how having a long road ahead makes this not a good starting point?