Google’s Translatotron is the latest system for the communication platform that’s infused with AI technology. The key giveaway with Translatotron is that the AI tech in the system is capable of detecting the tone of the words as well. Its core functionality remains to translate, but now, it will even tell you the tone of the speaker.
Although Google researchers say that Translatotron will be coming to commercial products, it might take ages. Right now, Translatotron isn’t as seamless as you believe, even though the conceptual idea is quite impressive. What you need to value here is Google that is translating the audio file directly into an audio file. It is skipping the part where the text is first translated into text. Now, that’s quite impressive!
Known as the end-to-end system by AI engineers, Google Translatotron is producing faster results as well as reducing the error count. Google says the model is processing isn’t raw audio from the data. Instead, Google is using something called the spectrogram data that is basically a detailed visualization of the sound produced. In absolute layman’s terms, this means that the language spoken is translated using pictures. Unbelievable, right?
AirAsia collaborating with Google Cloud to provide digital skills training in Malaysia
Google delays 30% cut policy to April 2020 exclusively for India
Indian startups to collaborate for an indigenous app store to fight Google monopoly
APAC Retailers Can Now List Their Products on Google For Free
Google follows Apple, enforcing stricter Play Store payment rules in India
© 2020 CIO Bulletin. All rights reserved.