Researchers at Microsoft have created software that can translate a user’s voice into a foreign languages, a solution with the potential to revolutionize communications, MIT’s Technology Review reported on Monday. “We will be able to do quite a few scenario applications,” said Microsoft researcher Frank Soong. “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.” Read on for more.
Along with removing the language barrier for travelers, the software could be used to help students learn a new language. Having the ability to hear one’s own voice could potentially make it a easier to imitate a new dialect. The technology could also be used to improve navigation devices, allowing a digital English voice to read text written in different languages.
The software currently has the ability to convert between 26 different languages — including Mandarin Chinese, English, Spanish and Italian — and it takes roughly an hour of training to develop a model that is capable of reading text in a user’s voice. The model is then converted, with the help of text-to-speech software, into one that can read text in another language. Sounds from the first model are then carefully tweaked to give the new text-to-speech model the full ability to sound out phrases in a new language.