Google creates technology to improve speech recognition

It gives the series “Here comes Google again, here, people!

I think the voice recognition is very good. I wonder how it will be better.

Google-voice-recognition-620x270A team of speech recognition-Google has announced that the company has been able to improve the acoustic models of its neural network, allowing voice commands to be recognised as the most efficient.

Thanks to a feature that the company calls “a Sort of a Temporal Conexionista”, and a sequence of techniques, and training, any discriminatory. The result is very simple: if you try to make a search for a speech, from now on, Google will understand it more quickly, more accurately and with less interference from the noisy environment, the company said.

The team says that the main difference is that the new technology uses the entire sentence is spoken, and not just the individual fragments of words, in order to identify what is being said. The technology of artificial neural networks, recurrent enables you to identify the sounds of speech in context.

The news is now reliant on multiple places, allowing you to better recognize your voice commands and searches for the app on Android and iOS. It is also present in the dictation in real-time.

Source: Look At The Digital

Translated by Yandex.Translate and Global Translator


Leave a reply cancel reply

This site uses Akismet to reduce spam. Learn to like your comments are being processed.