Google launched an update to Android’s voice search capability on Tuesday that incorporates speaker recognition to improve results.
The new “personalized recognition” feature is currently quite limited: it’s only available in English in the United States on Android devices running 2.2 and above.
When turned on, the new feature stores your voice searches and builds a statistical profile based upon the pitch and timbre of your voice and the pronunciation and pace of your speech.
“Speech recognition is based on statistical modeling,” Google’s help file on the technology says. “To recognize spoken words, we compare the input search to a statistical model of the language and try to find the closest match – the system’s best guess at what the user said…Knowing what you said in the past allows us to build specialized models that are designed to match your voice and your words.”
The feature then, over time, improves the speech recognition capabilities just for you. The recordings of your voice are associated to your Google Account and are not available to anyone’s searches but your own.
Personalized recognition can be enabled or disabled at any time through the Google voice recognition settings in your phone (Menu>Settings>Voice Input & Output>Voice Recognizer Settings), and all of your voice recordings supposedly can be disassociated from your Google account in your Dashboard under “Speech,” but since I’ve enabled the feature, that option has not shown up yet.