NTT Docomo’s Simultaneous Speech Interpretation System

The Japanese telecom mavericks, NTT Docomo recently demonstrated the real time speech-to-speech translation system. This technology, which promises to revolutionize the way we speak, was showcased at Wireless Japan 2011. Docomo, as always, is ahead of its competitors in putting to use novel technologies for the consumer benefit. For instance, the speech-to-speech translation system is completely based on cloud computing. This innovative use of applications, which are already available in the cloud means that the new product from Docomo is going to be exciting.

#-Link-Snipped-#

The simultaneous speech interpretation system development is still in initial stage. In the demo, real time speech translation was realized with a good level of accuracy. The demo included the reading of a newspaper by a staff member from the company while talking on the phone to a colleague. The person on the other side could hear the translated version of the conversation clearly. This feat could be achieved by a process, which includes a number of steps being executed sequentially at a very fast speed. Firstly, an application recognizes the words you speak. Then, these recognized words are translated in the required language. Now comes the exciting and the most complex part. The translated words are fused to form as real as possible natural speech after which you can hear the conversation.

The company admits that the system is still not foolproof as it uses a large number of services available in the cloud that it thinks is best as of now. The development of more tools that are sophisticated and their availability in the cloud will possibly improve the speech translation system. How much time this could take, nobody knows! Even then, if you are a Docomo user and think that you want to try this system, keep your fingers crossed. Docomo plans to launch the service for select (willing) customers who would give their feedback to the company even though translation is still not perfect. The problem persists in the voice recognition and translation stages of the complete translation process and the company is trying its best to come up with a better alternative service for the same.

The simultaneous speech-to-speech translation system can prove to be very useful for people who travel a lot and interact with people in different parts of the word on a regular basis. In addition, this could mean more business for a person without the need to learn different foreign languages. However, the most thrilling avenue that opens up is how people are going to interact with each other.  The fact no one can deny is humans are meant for interacting with each other through a very powerful tool: LANGUAGE. Therefore, if a person wants to talk to a distant friend, language should not be a problem. So is the launch of this system going to undermine the importance of languages? I do not feel so. On the other hand, the linguists have more responsibility on their shoulders. The responsibility is to make the languages more “modern.” In addition, they will now have to compile vocabulary in various lingos of the same language more accurately.

Check out the following video:



Source: #-Link-Snipped-#