December 14, 2010 | by Andrew Kameka
Google announced a short while ago that Voice Search is being fine-tuned for individuals. Though Voice Search for Android originally launched with “speech models broad enough to accommodate a wide variety of people,” Google is target search that better suits the individual.
Personalized recognition increases accuracy for the user by building a speech model on that person’s pronounciation. So users who can’t shake their southern twang or New Yawk accent can associate recordings with words that will produce more accurate results the next time they attempt to perform a voice search. The feature is currently available to English speakers in the U.S. on Android 2.2 or higher.
Google also promises that the new engine also adds better name recognition and faster results. I welcome all of these changes because I have experienced some problems with Voice Search here and there. Try as I might, my second-hand Jamaican accent sometimes makes “th” take on a “d” sound, so now I can customize Voice Search to know what I meant when such slip-ups occur.
Get more information on the set-up process at Google Mobile blog.