NOV 21, 2012
By Ofer Tirosh
You may be an avid Apple fan, but Microsoft recently contributed something truly amazing to the language and translation industry. Their new speech recognition tool translates spoken word into another language within a matter of seconds. It sounds like something straight out of science fiction, but it is now science reality and possibly the future of the translation industry.
A YouTube video has surfaced starring the Microsoft chief research officer, Rick Rashid, in which he demonstrates the new technology. In the video he speaks a string of English which is then translated automatically by the computer into a computer generated version of his voice speaking Mandarin Chinese. The video can be seen here; http://www.youtube.com/watch?v=5W5QgcQoKLM .Rashid’s demonstration takes place in front of a live audience and they go wild each time a sentence is translated.
Microsoft improved the existing speech recognition error rate by about 15% with this language software, making them leaps and bounds ahead of any competition at the moment. But how does this translation software work exactly? To put it simply, once the software picks up on the voice recognition, it uses translation software to convert the text into another language. Then speech synthesis is used to create the output or the words spoken in another language.
While this translation tool certainly isn’t error proof, it’s still pretty cool and far and away better than those days of using free online translation software. Eat your heart our apple lovers, because this is a serious win for Microsoft.