Google has upped the ante for its voice search by making the feature a lot more personal.
The Mountain View, Calif. company has taken its voice recognition service and personalized it for users. When Google originally unveiled the service on its Android phones two years ago, it had the ability to accommodate a variety of people regardless of age, gender or accents. With this latest update, Google voice search can recognize individual people.
Google says this speech model will allow the company to deliver greater recognition accuracy. A spokesperson of the company said it uses statistical modeling to create this personalized speech recognition service.
To recognize spoken words, we compare the input speech to a statistical model of the language and try to find the closest match -- the system's best guess at what the user said. The statistical model is huge -- it must cover all of the fundamental sounds of the language (phonemes), all of the words, and all of the different ways that the words can be strung together in the spoken language, the spokesperson said in an email.
Furthermore, it must capture all of the variations among users that happen when a language is spoken, for example all of the different dialects and accents and individual differences in the sound of the voice.
The spokesperson said over time, the system will better recognize the acoustic and language characteristics of the individual. These characteristics ultimately are turned into a voice profile, a series of numbers that tells the computers how the user is different from everybody else.
For now, the speech recognition service is only available on Android 2.2 operating system based phones or higher. The company said there were no immediate plans to bring it to any other platforms. The spolesperson said voice recognition is critical to the company's plans in delivering the Android platform.
Voice recognition is important to Google as a whole. All of the world's information is spoken, and if our mission is to organize the world's information, it needs to include the world's spoken information, the spokesperson said.