NUANCE ACQUIRES PHILIPS SPEECH RECOGNITION SYSTEMS

Document Sample
NUANCE ACQUIRES PHILIPS SPEECH RECOGNITION SYSTEMS
> > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S > > N E W S > > T R E N D S > > A N A LY S I S









University Project Could Yield a eventually be used in real trials at The

Boston Home, a nursing home in



New Mobile Speech Application Dorchester, Mass.

“At present we have a prototype

that we’re deploying in the lab, and we

Professor Nicholas Roy and his colleagues at the Massachusetts do hope to begin trials with real users

Institute of Technology are working to develop a self-navigating within the next couple of years at The

Boston Home,” Teller says. “There’s a

wheelchair capable of learning different locations within a lot of interesting stuff that happens

building and then transporting users did you mean? It initiates a dialogue when you try to roll out a system like

to those locations in response to verbal with the user rather than simply tak- this to a real community. Most, if not

commands. ing speech in stone and trying to do all, of the residents at The Boston

Through speech recognition tech- the best it can with it.” Home have some sort of neuromuscu-

nology, a user will be able to speak Instead of relying on manually cap- lar disability…and in many cases that

commands—Take me to the study or Go tured maps, the MIT wheelchair slurs their speech, makes it soft, and

to the conservatory—and the wheel- learns about its environment through that is going to compromise the accu-

chair will automatically maneuver a guided tour. The chair is pushed racy of the speech system.”

from place to place based on a map around its new environment as the But Teller says the wheelchair’s de-

stored in its memory. user identifies rooms and locations. sign—particularly its ability to ask users

“We’re using speech recognition de- The wheelchair prototype uses a WiFi clarifying questions—addresses those

veloped inside the Computer Science system and a network of WiFi nodes challenges. “Hopefully through repeated

and Artificial Intelligence Laboratory in each location to generate its maps interactions—just the way that a human

developed by the Spoken Language and navigate through environments. might initiate an interaction if he or she

Systems Group,” says Seth Teller, pro- “The other novel thing we’re doing is doesn’t understand what was just

fessor of computer science and engi- using speech to help the wheelchair said—the wheelchair can prompt the

neering and head of the Robotics, learn a map of the environment,” Teller user to try to get a

By registering with docstoc.com you agree to our
privacy policy and terms of service

Successfully added document to cart!

Successfully added document to cart!