American Sign Language Recognition using Kinect Skeleton features

This a video presentation of the ongoing effort in American Sign Language recognition using the Kinect by researchers at the College of Computing, Georgia Institute of Technology.

For inquiries contact:
ZAHOOR ZAFRULLA
(zahoor@gatech.edu)

Dr. Helene Brashear
(helene@gatech.edu)

Peter Presti
(peter.presti@imtc.gatech.edu)

Dr. Harley Hamilton
(hhamilton@gatech.edu)

Dr. Thad Starner
(thad@gatech.edu)

Important Links:

Copycat Game:
http://cats.gatech.edu/content/copycat

Signing Made Easy: NSF special report
http://www.nsf.gov/news/special_repor...

Georgia Tech Gesture Toolkit:
https://wiki.cc.gatech.edu/ccg/projec...

Hidden Markov Model Toolkit:
http://htk.eng.cam.ac.uk/

OpenNI framework
http://openni.org/

PrimeSense NITE Middleware
http://www.primesense.com/?p=515

• Category

  Science & Technology

[geekindy]

Great work, and a nice touch presenting the video without sound, given the intended user base.

[tangos]

This is neat. Maybe one day learning ASL will be a video game that anyone could try for.

[Dennis Wee]

there should be a translation about this so that educators that can speak can comprehend more.

[Aztecgames]

Don't want to spam your comments, but I wanted to get some deaf peoples opinions on my xblig title 'RumbleVoice'.

It allows you to speak into the headset on xbox and the controller will rumble to match the speech or sound.

I had thought this might be useful for someone who is deaf and trying to work on pronunciation. Have a friend say words normally to get used to the feel and then they could practice until they could better reproduce the feel and sound of the word.

Check it out or PM for a code