Google’s Voice Search was designed to be easy to use for as many people as possible when it launched. That meant that the parameters used for voice analysis had to be general rather than specific. Accents and gender were not as much a concern as the not-so-simple task of recognizing the human voice.
But today, Google announced–via the Google Mobile Blog–the availability of Voice search 2.1.1 in the Android Market, and it offers personalized voice recognition, meaning that you can train e software to recognize your specific speech patters by selecting words as representative. The option can be enabled in the voice recognition settings panel or via your Google Dashboard. Users are also prompted to opt-in after installing or updating Voice Search. Personalized speech recognition can be disabled at any time.
The service is currently available only for English speakers with Android 2.2 and up. The update also features performance optimizations.