Microsoft’s Universal Translator Translates Your Own Voice in Other Languages
Who needs months of lessons in foreign languages when one handy app can get you talking in Mandarin or French without an ounce of effort? Microsoft Research has developed new software that translates your speech into another language but with your own voice.
Researcher Frank Soong demonstrated the software last week at Microsoft's campus in Redmond, Washington at the company's TechFest 2012. Like other translators the software can translate a speaker's voice in any combination of 26 different languages but with one big improvement. Unlike other translators Microsoft's program uses the speaker's original voice in its translations not a computerized version of the speaker's voice. Even better the software creates a 3D image of the speaker's head to look like the person is actually speaking the foreign language.
"We will be able to do quite a few scenario applications," said Soong, who created the program with colleagues at Microsoft Research Asia in Beijing, China. "For a monolingual speaker traveling in a foreign country, we'll do speech recognition followed by translation, followed by the final text to speech output [in] a different language, but still in his own voice," said Soong.
A universal translator like the one in Star Trek has many uses. Students could use it to learn new languages. Travelers will have no problem understanding different languages abroad while entrepreneurs and politicians will feel secure that the true intent of their words will get understood by their foreign counterparts because the program uses the speaker's original voice in its translations. A person's speech patterns could convey more information that just what the person is saying.
"Preserving voice, preserving intonation, those things matter, and this project clearly knows that," Shrikanth Narayanan, a professor at the University of Southern California told MIT's Technology Review. "Our systems need to capture the expression a person is trying to convey, who they are, and how they're saying it."
The software still needs to spend an hour with the user learning the particulars of the person's individual speech patterns but once the software learns the speech patterns the translation happens in real time. Once this program is released anyone could install it into their smartphones or tablets and then use the software to speak to anyone in any part of the world.