Google Finds a new way to help its speech-impaired users

Google is one of those tech giants that constantly trying to evolve by bringing comforts in the life of its users with a variety of products every once in a while. With the updating tech products every year, back in may 2019, Google launched a neural network-powered algorithm to help its users with speech-to-speech translation. The core purpose of the Google team is to provide a product that makes its users rely more on it than any other platform.

By launching the new neural network algorithm Google tried to make its users more dependable but as unique and attractive this product sounds it also holds a drawback. The new algorithm was interesting enough for all customers but people with speech impairments were not much impressed with that.

Google’s new update to its users

The neural network algorithm by Google could easily do the speech-to-speech translation in real-time but it was inaccessible for speech impaired people and for that purpose Google’s artificial intelligence teams have been working on a new software to specifically help people with speech disorders.

Last week, Google's AI engineers announced a new speech-to-text technology aimed to help people with verbal disorders and dysarthric speech. Parrotron is the software that runs on end-to-end deep neural network and converts any impaired speech patterns to fluent speech. The most unique feature about this model is that it doesn’t depend on any lips movement to get cues.

Google’s help to make people with speech flaws talk freely



Google fed the neural network with “corpus that consists of millions of spoken word pairs” this new technology is specifically designed to reduce the word error rates for any deaf speaker from 89 percent to 25 percent. Google promises to bring new updates to this AI too. Researchers have also shared a variety of demo videos to the users to help them understand the technology easily.
"Parrotron makes it easier for users with atypical speech to talk to and be understood by other people and by speech interfaces, with its end-to-end speech conversion approach more likely to reproduce the user’s intended speech.", explained Fadi Biadsy, Research Scientist and Ron Weiss, Software Engineer, Google Research in a post on Google AI blog. Adding further, "More exciting applications of Parrotron are discussed in our paper and additional audio samples can be found on our github repository. If you would like to participate in this ongoing research, please fill out this short form and volunteer to record a set of phrases. We look forward to working with you!"

Bottom Line

The core focus of Google is always to bring comfort for its users, it always tries new technologies and products for its users. The new speech-to-speech model that works in real-time is a revolutionary AI that not only for fluent speakers but the new updated technology is a product to help people with speech impairments as well. This report further highlights the Google's efforts to bring new technology that could help its customers in all the ways possible.

No comments:

Post a Comment