Speech recognition has been widespread for over a decade now. Nuance, a leader in speech technology, brought their first commercial speech application to market in 1996. That’s only 13 years ago!
It’s been slow to break away from the ultra annoying call-steering customer service and PC dictation applications and onto something more widespread and useful, but this year marks a shift for Voice User Interfaces from relative obscurity to mainstream mania.
iPhone 3GS
Upgrades to the new iPhone 3GS operating system include the ability to select playlists for playback and make phone calls using voice commands. There is also text-to-speech for things such as asking your iPhone “What song is playing now?”. They have also opened this up to developers in the API so this new feature is available to 3rd party Apps.
Xbox 360 Project Natal
With the release of Project Natal, you can control your console and games using your voice. Check out the video from 2:58 mark onwards for a showcase of the voice features.
Before Natal, we had…The Clapper?
Conclusion
VUIs and Voice Recognition are still infants. They have to grow up and they haven’t even hit puberty. Recognition benefits from personal training, from getting familiar with the way you vocalize words. I hear the new Dragon Naturally Speaking Software is 98% accurate out of the box, but there’s still room for improvement and I doubt it is that accurate when you are talking about a niche subject that has many lingo terms. Imagine we had a store of speech data that we could point the software to so that we don’t need to personalize it? This would help put your ramblings in context and improve accuracy.
Here is a shortlist of other things using VUIs:
Nintendo DS
Ford vehicles - Microsoft Sync
Onstar
411 and directory services
Self serve customer service
Blackberry (Powered by Nuance)
Windows mobile
Spinvox visual voice mail
Jaws




