Google is now able to hear the user even in noisy environments
The company Google has announced improvements to its voice recognition system And the fact is that voice searches are a recurring function for users mobile , in addition to the future or already present of communication with devices such as smart watches Something in which Google has always stood out for having one of the finest and most precise recognitions in the world, recognizing what the user says almost immediate and with fewer failures than the competitionA matter that is now even better and faster thanks to your advances.
The advances come with an improvement of what they call neural network of acoustic models And now these models of Google are supported by Recurrent Neural Networks (RNN in English). That is, a system that analyzes what is dictated out loud by the user, taking into account the sounds being analyzed and which was the previous sound All this with analysis methods that significantly improve the precision of this recognition system. In addition, the company has improved this system by avoiding analyzing sound prediction data at every moment. All this introducing noise and reverberations into training these speech recognition models. In other words, they have managed to improve their technology to make it more precise, perfectly understanding what the user has said, faster than before, and even when the user is in noisy environments and poor acoustics
In this way, Google urges users to try voice searches through their application of the same name for both Android and iOS Y is that the new models for recognition have already been implemented in this search tool Something that, in the case of terminal users Android, you can also use dictation mode, by tapping on the microphone icon anywhere you can type.
With this, the search experience and voice interaction with this application is much more agile, avoiding waiting times that Before, they were used to analyze what the user had dictated and search for it on the Internet. In addition, the user has to worry less and less about articulating each word and phrase robotically, being able to speak naturally with the search application to request what they want find without too much effort, just as if you were telling someone else.All this seasoned with useful issues such as noise detection and isolation A really important point to make use of this tool in any environment, without fear of being misunderstood or fail the search. The latter becomes especially important through smart watches or even the car dashboard Devices where Android is rolling out and you will enjoy a smoother user experience in any situation.
At the moment the new models and voice recognition technology are available through the application Google that can be downloaded free via Google Play and App Store .
