Microsoft’s CEO recently demonstrated live speech-to-speech translation at a presentation in China. The speech recognition software converts spoken word into text, then translates that text into the desired language while correcting the word order so that an accurate sentence is delivered in the target language. Finally, the sentence is spoken out loud.
The most exciting part? The sentence is spoken out loud, not in a computerized voice, but in the speaker’s voice. The software is able to replicate the speakers voice and vocal patterns in order to create a seamless speech-to-speech translation. The software, namely the translation accuracy, is still in the developing stages, but the demonstration was a taste of what’s to come with advances in technology.