Google has recently one of their latest developments: Translatoron. What does it do exactly? It actually takes your voice and speaks in another language for you.
In more technical terms, it’s known as a speech-to-speech translation system. And although it’s not exactly the first time such a thing has been done, this system created by Google has one very unique thing about it, and that’s the fact that it can actually retain the voice of the original speaker.
What this means is that this innovative technology could actually make it sounds as though you are speaking a language that you don’t even know. This is completely groundbreaking for the global issue of language barriers between people around the world, and it’s all thanks to artificial intelligence.
Based on Google’s blog – most of these kinds of systems follow three steps. The first being transcribing the speech, then translating into the language of choice. Then once that’s done, the translated speech is turned into audio.
What’s different about Google’s system, Translatoron, however, is that it skips over this standard process, and simply takes the speech and turns it into an image that gives off audio frequencies, known as a spectrogram.
As it stands, the system is far from perfect and has yet to be released onto the market, however, the prospect of it is obviously exciting and the future of communication is looking brighter than ever.