Near-Instant Universal Translator Taking Shape

Via UWR.org

How close are we to breaking down global language barriers? Microsoft researchers demonstrate that we’re closer than anyone could’ve imagined.

For the last 60 years, computer scientists have been working tirelessly to build systems that can genuinely understand what a person says when they talk. The most primitive of techniques involved simple pattern matching, which was flawed by the fact that various people said the same things differently, and even one particular individual could enunciate or speak the same phrase in a multitude of ways.

We saw modest improvements over the years, but nothing truly significant until two years ago, when Microsoft engineers working with the University of Toronto implemented a technique called Deep Neural Networks that more dramatically mimicked the speech recognition and translation process the human brain goes through.

Though the technology is not yet commercially available, the improvements demonstrated in the video below are dramatic.

Speech Recognition and Translation

To put things in perspective, the speech recognition technologies our smart phones are currently outfitted with experience an error once every 4-5 words. In the new, still unpolished deep neural network technology Microsoft is introducing, they’ve reduced that error rate to once every 7-8 words. While this is a monumental feat by itself, it’s what they’re doing with these new-found abilities that could have amazing worldwide humanitarian implications.

With Microsoft’s new ‘universal translator’ of sorts, not only can an individual’s speech instantly be recognized, it’s translated in near real time, rearranged in a way that makes sense to native speakers, and then relayed in your own voice for familiarity and comfort. Not only is your tone translated, but your cadence is replicated, which is essential to conveying critical points in human speech and avoiding misunderstandings. The below speech by Microsoft’s Rick Rashid showcases the future technology to 2,000 students in Tianjin, China.

If Microsoft can get a working ‘instant translator’ to market, especially if it’s bundled into their phones or tablets, it’ll go a long way towards re-establishing the creative and innovative rapport with consumers they’ve sorely lacked over the last decade.

Microsoft via New Scientist

Comments

comments