Google’s prototype AI translator translates your tone as well as your words

We all know that communication depends on something more than what you say. How you say it is often equally important. That's why the latest AI translator for Google prototypes not only translates the words that come out of your mouth, but also the tone and cadence of your voice.

The system is called Translatotron, and Google researchers explain in detail how it works. a recent blog post They do not say that Translatotron will come to commercial products soon, but that will probably happen in time. As Google's translation chief explained to The Verge earlier this year, the company's goal at this time is to add more nuances to its translation tools, creating a more realistic discourse.

You can hear how this sounds in the audio samples below. The first clip is the entrance; The second is the basic translation. and the third tries to capture the voice of the original speaker.

Entry (Spanish)
Translation of translatotron
Translatotron translation with inflection

As you can hear, it is not a perfect translation, but it is still impressive. You can listen to many more Translatotron audio samples here.

While capturing the inflection of a speaker's voice is what most impresses laymen, the attraction of Translatotron for AI engineers is that it translates the voice directly from the audio input to the audio output without translating it into the usual intermediate text.

This type of AI model is known as an end-to-end system, because there are no stops for subsidiary tasks or actions. Google says that doing end-to-end translation produces results faster and avoids the risk of introducing errors during several translation steps.

Perhaps even more interesting, the data that the model is processing is not raw audio. Instead, it uses spectrogram data or detailed sound visualizations. In essence, that means that we are translating speech from one language to another using images, which is amazing.

As always with Google's translation efforts, there are reasons to be skeptical about how systems of this type will work in nature. The company often introduces new and ambitious speech and translation tools, and often performs less fluently than we expect. Even so: the future advances, and AI's translation is only improving.

Please Note: This content is provided and hosted by a 3rd party server. Sometimes these servers may include advertisements. igetintopc.com does not host or upload this material and is not responsible for the content.


Open

Close