Mountain View, California — Google has possibly done it again! Rumors on the street are abuzz that the search titan is planning to break the language barrier and Google engineers are working on a translator for Google Android smartphones to convert one language into another language on-the-fly quickly enough to allow speakers without a common language to communicate with one another in near real time.
According to reports appearing in The Times Online states that the Internet giant has been developing new translation software for mobile phones that will see two people from different sides of the world able to converse in real time, without needing to speak the same language.
“We think speech-to-speech translation should be practicable and work reasonably well in a few years’ time,” Franz Och, Google’s head of translation services, told the Times Online on February 7.
The concept behind Google’s plan is simple: when a user speaking one language says something into a phone, Google’s translator software, which would be running on the recipient’s phone, would interpret what the person said, translate it into the recipient’s language, and repeat it back in that second language. In order to limit conversational pauses, the company apparently plans to translate as phrases are spoken, rather than waiting until each sentence is completed.
“Everyone has a different voice, accent and pitch,” explained Och. “But recognition should be effective with mobile phones because by nature they are personal to you. The phone should get a feel for your voice from past voice search queries, for example.”
In essence, achieving real-time digital translation will be no easy feat. Google anticipate to make the application available “in a couple years.”
Voice-recognition technology has traditionally struggled to work quickly and accurately. Now, seems to be a few problems with this concept in the past that have surfaced when trying to do anything involving voice translation.
First off, the pronunciation problem. Many applications built into current mobile devices require rigid commands to function. Everybody has different ways of saying things and often times it does not translate correctly. For example, in order to initiate a call using my Bluetooth headset, I’m required to clearly say “call” followed by the person’s name. I have got to speak like a robot and even when I do, the software sometimes fails to understand my commands.
However, Google Voice, the search leader’s telephone service, which also faces difficulty in making exact translation of the speech to text. Whenever a caller leaves a voicemail on a Google Voice user’s answering service, Google delivers a transcript of the message to the user’s voice inbox, which is identical to an e-mail inbox. Many users have found that those transcripts are rife with errors.
Google has its work smoothly sorted out for it. Translation presents a tougher challenge than re-imagining e-mail or copying Microsoft Office as a cloud-based service. Humans are nuanced communicators — which is why, to date, the translators readily available are basically elaborate gimmicks, limited by the size of their dictionaries and inability to parse phrases.
As usual, Google’s goal is high-minded: to distribute real-time translation of spoken meaning, rather than just words. To carry out this, the company is sewing together its voice recognition, 52-language text translation, and text-to-speech technologies into an integrated voice-to-voice translator.
“Clearly, for it to work smoothly, you need a combination of high-accuracy machine translation and high-accuracy voice recognition, and that’s what we’re working on,” Och said.
Google, however, is not the only one trying to bring out real-time voice-to-voice translation software. In June 2009 a company called Sakhr announced it was pioneering and developing “the world’s richest knowledge base for Arabic natural language processing.”
Sakhr has already collected a list of high-profile companies (including the US Departments of Defense, Homeland Security and Justice) who have started to put its instant translation service to the test on the Blackberry and iPhone mobile phone platforms.