Microsoft demonstrated new translation software at their TechFest 2012 event, which took place last week, revealing some unique features of their Monolingual TTS translator.
The first unique feature is that, while users talk into the translator as normal, the Monolingual TTS system outputs the translates speech by generating facsimile of the user’s voice rather than a using a computer-generated voice. The translator also generates a 3-dimensional copy of the speaker’s head which repeats the words back.
“The ambition is to create a visualized language teacher that can be engaged in many aspects of language learning from detailed pronunciation training to conversational practice. An initial implementation is a photo-realistic talking head for pronunciation training by demonstrating highly precise lip-sync animation for any arbitrary text input.”
The translation system still takes some training in order to get the nuances of a given speaker down. Typically in the region of an hour’s worth of speaking is needed before one can use Monolingual TTS effectively. The translator recognises 26 languages at present.