Microsoft has demonstrated new software that can pull together real-time multilingual vocal translations using your own voice. Monolingual TTS currently handles 26 different languages, although it’s not instant just yet — it takes about an hour of training to get the experimental software acquainted with your own utterances. Demonstrated at Microsoft’s TechFest 2012 showcase, the software can even mix up foreign language pronunciation of place names with directions in your native tongue. It also complements those efforts with a 3D image of your head, animating your lips along to the foreign words you’d otherwise butcher. See how an algorithm-educated floating head handles Mandarin — and how it’s all done — right after the break.
Microsoft demos vocal translator at TechFest 2012, uses your own dulcet tones (video) originally appeared on Engadget on Mon, 12 Mar 2012 08:48:00 EDT. Please see our terms for use of feeds.
Permalink 
PCWorld | 
Microsoft Research | Email this | Comments