Certainly an April fool's joke... however, work like this has practical applications for patients whose physical disabilities prevent them from using the traditional computer inputs. I did some work with mapping shoulder movements to control a powered wheelchair at the Rehabilitation Institute of Chicago during an internship in 2007, and I had the impression that there is a lot of room for improvement in current devices. One interface commonly used today is the sip-puff switch:
http://www.techready.co.uk/Assistive-Technology/Sip-Puff-Swi...
I wouldn’t be surprised if there were people doing serious work in developing facial gesture UIs.
Users visiting Web sites that contain adult content sometimes make unconscious facial expressions. If Opera keeps opening Speed Dial and Zooming In and Out, please be aware that this not a bug. It may be useful to disable Face Gestures for a better experience. Also remember that wiping your cache is recommend after visiting untrusted websites."
"By using an internal technology dubbed Face Observation Opera Language, we are able to recognize pre-determined facial expressions and match them to commands on the Opera browser."
and you can see that they have already realized realtime face tracking with dense registration ( which allows expression recognition ) based on the LK-inverse composition algorithm applied to AAMs
10 comments
[ 3.7 ms ] story [ 32.8 ms ] threadUsers visiting Web sites that contain adult content sometimes make unconscious facial expressions. If Opera keeps opening Speed Dial and Zooming In and Out, please be aware that this not a bug. It may be useful to disable Face Gestures for a better experience. Also remember that wiping your cache is recommend after visiting untrusted websites."
Tweeted it. Went back and read it properly. Realised it was april 1st. Do'h
Face Observation Opera Language, eh?
Look at some of the works done by Iain Matthews and Simon Baker at CMU here:
http://www.ri.cmu.edu/research_project_detail.html?project_i...
and you can see that they have already realized realtime face tracking with dense registration ( which allows expression recognition ) based on the LK-inverse composition algorithm applied to AAMs
Facinating stuff =)