27 Mar ,2020
The present of machine translation
With the increasingly frequent global interconnection, almost everyone is eager for the real-time translation inverse technology to become a reality soon. With the unremitting efforts of this generation of scientists, science fiction is stepping into reality.
In order to realize the dream of machine translation, foreign scientific research institutions and enterprises represented by IBM, Google, and Microsoft have set up machine translation teams one after another to specialize in intelligent translation research and aim to break language barriers.
IBM Machine Translation
IBM has been conducting large-scale research in this field since 2001, and has made major investments in the field of translation between English, Arabic and Chinese.
In September 2009, IBM officially launched ViaVoice Translator machine translation software, which laid the foundation for automated translation. After this, the research project undertaken by IBM did not make any significant progress and was subsequently abandoned.
Since 2011, with the rapid development of speech recognition, machine translation technology, DNN (Deep Neural Network) technology and the needs of economic globalization, the research of automatic spoken translation has become a new research hotspot in the field of information processing.
Google Machine Translation
Google officially launched an upgraded version of machine translation services on its Android system in January 2011. Now, Google Translate can translate between more than 70 languages. In 2012, Google Translate was used by approximately 200 million users.
So far, this service is free. But for an automated translation service, multinational companies will be happy to use this service, which will bring high value to it, which means that Google will almost certainly be able to generate a lot of revenue through this translation tool in the future.
Microsoft Machine Translation
In 2012, Rick Rashid, the global head of Microsoft Research, announced the success of the machine translation project at a conference in China. He was speaking in English at the time, and a Microsoft computer translated what he said into Mandarin at the same time, and spoke it in Rashid's own voice. From this point of view, this translation program by Microsoft has become infected with a touch of human .
Microsoft's Skype announced in December 2014 a preview of real-time machine translation, real-time translation in English and Spanish, and real-time translation of text in more than 40 languages.
The most important technology behind Skype translation is the huge improvement of speech recognition. The combination of Microsoft neural network speech recognition technology and machine translation allows users to translate better in one-to-one different languages.
The future of machine translation
As machine translation technology advances, another worry is creeping in: will machines grab the job of simultaneous interpretation? Will people in this industry lose their jobs?
It is too early to worry about this issue. Although it is now possible to use machines to simulate human intelligent activities and simulate human understanding of semantics through some practical technologies, there is still a long way to go before thorough artificial intelligence and semantic understanding.
If machines can really achieve accurate real-time translation one day, their role is more likely to assist workers in related industries to work more easily and conveniently, rather than replace them.