Microsoft Is Getting Close To Perfecting A Universal Communicator

H

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
Time has posted a somewhat lengthy article on the status of Microsoft's Skype Translator and perfecting a universal communication tool.

An audience of several hundred reporters and industry insiders watched on as Pall and a native German speaker held a nearly flawless conversation through the company’s prototype of Skype Translator. Roughly a second after Pall Spoke, subtitles in German and English appeared at the bottom of the screen, and a synthetic Siri-like voice read the words aloud to the German caller.
Click to expand...
 
Been using skype translator and it's nice, but the key thing I need is translating group chats which it doesn't do yet, would be nice if they had google's language detection and translated instead of only working one to one when you specify which language they'll use.
 
It may be a real time translator, but universal it is not.


For it to be universal it would need logic to automatically figure out and translate any language, including completely unknown languages it has never been exposed to before.
 
I was at GTC 2015 last week, and these types of advancements and many others were at the focus of the entire conference. It was pretty amazing to see some of the ways Deep Neural Networks, Convolutional Neural Networks and others, and even combinations of were being used for speech, image and even behavior recognition.
 
I can't read the article right now, but is this really more than something like this:

speaker -> voice to text -> bing translator -> text to speech -> listener

In which case this isn't amazing at all. All of these technologies are old hat, and stringing them together in a row doesn't impress me much.

They will likely result in the same kind of useless translations as using google translate to read that foreign web site with "leaked benchmarks"...
 
Zarathustra[H];1041504033 said:
I can't read the article right now, but is this really more than something like this:

speaker -> voice to text -> bing translator -> text to speech -> listener

In which case this isn't amazing at all. All of these technologies are old hat, and stringing them together in a row doesn't impress me much.

They will likely result in the same kind of useless translations as using google translate to read that foreign web site with "leaked benchmarks"...
Click to expand...
I read the article. They don't focus on the components in use, just that they have made quite a bit of progress in regards to understanding spoken words. Second and fourth components are probably what all the fuss is about. It's interesting stuff. Hopefully the tech makes its way to Cortana (I have a feeling it will).

I doubt that it will likely result in anything similar to google. Using Cortana right now on my Windows Phone is a better experience than on my android based LG G3. Based on the article and my experience with the current technology, this should be something to look forward to.
 
Knowing Microsoft they will have to listen to ads in both languages first. :D
 
You must log in or register to reply here.
Back
Top