Microsoft software translates speech, into your voice

There may soon be a way to speak 26 languages without stepping foot inside a classroom or cracking open a textbook. Researchers at Microsoft have found a way to learn the sound of your voice and use that in conjunction with machine translation to make the translated audio sound just like you!

In theory, the software studies the sound of your voice to be able to convert it to the language of your choice. It first uses speech recognition, then translates the text, and then plays it back in the language of your choice. Click on the link below to hear example audio from this software.

Why is this so exciting?

The improvement of speech-to-speech translation technology has helped smartphone users travel more easily, without the use of heavy phrase books and dictionaries.

While machine translation is still not extremely accurate, this is an interesting development for those who want to use speech translation for recreational purposes. I wouldn’t recommend this for people who need interpretation for medical, legal and business purposes however.

All bias aside—we are a translation company after all—machine translation should be used sparingly for any reason other than recreation and research, where the quality of the translation doesn’t matter.

But as we always say, in critical situations (medical, legal, etc.), human interpretation is key. Nothing can give higher quality than a certified human interpreter.

The new technology announced by Microsoft is a step toward a future where country lines are blurred, where travel becomes easier and different cultures can be explored more easily!

Read the full article and listen to samples of this technology: Software Translates Your Voice into Another Language