Microsoft Research is working to develop a virtual voice translator generated from a real mode that is responsible for imitating the tone and timbre of users when translating the content into 26 Languages.

Subtitles have their days numbered by virtue of the work you're doing Microsoft with its real-time multilingual translation system, A technology that allows you to simulate conversations in 26 Different languages through a virtual avatar generated from a real model.

This Microsoft project captures intonation to make translations more real, allowing the user to say a sentence in their language and the translator to respond in another with the same tone of voice. To do this, the person in question will have to speak through a webcam or a Kinect sensor for a 2D synthesizer and a 3D modeling program to generate the virtual character that will give life to the interactive translator. In this way the character will be able to perfectly gesticulate any type of pronunciation, being accompanied by a somewhat less robotic voice than we are used to hearing in automatic transcription systems.

Apparently, the Microsoft team has done a test of the system in Redmond, where they translated sentences into Spanish, Italian and Mandarin.

Watch Video


You liked this article?

Subscribe to our RSS feed And you won't miss anything.

Other articles on
by • 16 sea, 2012
• section: Infrastructure