Follow this blog:
RSS

Ford, Nuance: Voice recognition will recognize intent

By | June 22, 2011, 3:32 AM PDT

Ford and Nuance Communications, a leading voice recognition software company, said they will collaborate to make the automaker’s SYNC in-car entertainment system recognize intent.

If successful, your car will ultimately interpret what drivers are looking to do. These voice capabilities could understand intent and meaning of words that drivers might say. SYNC would then coach you on how to complete tasks in the car.

SYNC already understands 10,000 commands, up from 100 in the first generation system. But to truly understand intentions, SYNC will need to understand more commands.

Nuance and Ford said they are testing “a number of new algorithms” to divine intent and meaning. These algorithms are designed to flag common words and phrases that are frequent in conversation, but not commands in the SYNC system. These new algorithms are designed to bridge the gap and coach the driver to issue the right commands.

These intent algorithms will interpret a phrase like “I wanna call John Smith” and translate it to “Call John Smith.” If there’s not enough information from the driver, SYNC will ask more questions.

This process, which will be an improvement, still has a problem: SYNC will have to coach the driver to use its vocabulary. Ultimately, SYNC’s system will have to adapt to how humans actually speak. Nuance has several language models for SYNC.

Ford said that 85 percent of its car owners are using voice controls, up from 70 percent in 2010.

Separately, Ford said that it is rolling out its AppLink program to 10 more vehicles in 2012. It also has 2,500 developers creating in-car applications.

Start your week smarter with our weekly e-mail newsletter. It's your cheat sheet for good ideas. Get it.

Larry Dignan

About Larry Dignan

Larry Dignan is the editor-in-chief of SmartPlanet.

Larry Dignan

Larry Dignan

Editor-in-Chief

Larry Dignan is editor-in-chief of SmartPlanet and ZDNet. He is also editorial director of TechRepublic. Previously, he was an editor at eWeek, Baseline and CNET News. He has written for WallStreetWeek.com, Inter@ctive Week, New York Times and Financial Planning. He holds degrees from the Columbia University Graduate School of Journalism and the University of Delaware. He is based in New York but resides in Pennsylvania.

Follow him on Twitter.

Larry Dignan

Larry Dignan
Larry Dignan does not hold any investments in the companies he covers.
If you liked this, don't miss...
2
Comments

Join the conversation!

Follow via:
RSS
0 Votes
+ -
Sorry - were you talking to me ... ?
My other and I have been (unwittingly) testing out just such a user interface on the assumption that the software can become as good as I am at deciphering her meaning. If I understand her correctly (!) she does not consider this a particularly challenging assumption.

I really do hope Ford are also developing a device for keeping the speaker seriously on topic. When she is multi-tasking (should that be multi-speaking?) the right word can have the wrong meaning, sometimes with consequences out of all proportion to the misunderstanding.
Posted by PassingWind
22nd Jun 2011
0 Votes
+ -
thanks for sharing
Great!!! thanks for sharing this information to us!
sesli chat sesli sohbet
Posted by yarinsiz
Updated - 26th Aug 2011
Join the conversation
Formatting +
BB Codes - Note: HTML is not supported in forums
  • [b] Bold [/b]
  • [i] Italic [/i]
  • [u] Underline [/u]
  • [s] Strikethrough [/s]
  • [q] "Quote" [/q]
  • [ol][*] 1. Ordered List [/ol]
  • [ul][*] · Unordered List [/ul]
  • [pre] Preformat [/pre]
  • [quote] "Blockquote" [/quote]

Join the SmartPlanet community and join the conversation! Signing up is fast and free. Don't wait -- we want to hear your opinion!