IT and Society: Japan unveils real-time speech translation for mobile phones

Language barriers may soon be removed with the help of new real-time speech-to-speech translation technology for mobile phones now in development.

The technology, which was presented by NTT Docomo, Japan’s mobile phone operator, at the Wireless Japan 2011 exhibition (May 25–27, Tokyo) uses a mixture of cloud-based services. Using the cloud specialists have managed to combine speech recognition and synthesis, and machine translation.

Demoed at Wireless Japan 2011, one person on the DoCoMo staff read a newspaper article in Japanese at the research center in Yokosuka while the person on the other end of the phone call received the real-time translation at the exhibition hall.

Atsushi Sato, the company’s representative, said the voice recognition isn’t 100 percent yet. He said that for customers to use it, DoCoMo will have to improve both recognition accuracy and translation accuracy. However he believes that even though the technology is still not perfect, customers would be able to appreciate its convenience. Therefore already this year NTT Docomo is intending to implement the technology in the test mode by recruiting volunteers, who would use it and provide feedback.