Real-time language translation device whispers into your ear in real time as it’s listening to the person speaking

Tuesday, September 12, 2017 by

Traveling abroad? Nervous that you will find yourself in a place where no one speaks your language? You might find yourself quickly becoming grateful for a new artificial intelligence device; an earpiece called  Translate One2One that can translate spoken conversations in a matter of seconds.

Translate One2One, which was set to launch in international markets by July, currently works for eight languages: French, German, English, Mandarin Chinese, Japanese, Italian, Brazilian Portuguese, and Spanish.

West Gosford, Australia-based startup Lingmo International, the company that designed the earpiece, said the device, which was introduced on the first week of June at the United Nations Artificial Intelligence for Good Summit in Geneva, Switzerland, is able to get past common translation mistakes. The built-in microphone analyzes spoken phrases, which is then translated to another language within three to five seconds.

Lingmo founder Danny May, who is a former plumber and the brains behind the product, said: “On average we’re going for 85 percent accuracy. We’re moving away from word-for-word translations and taking [phrases] in context — and currently 85 percent accuracy is what we’re getting from doing field testing. We’re taking up to a 30-second block. We really focused on the speech recognition side — for the simple reason that if the speech isn’t picked up correctly, then the translation will come back incorrect.”

Translate One2One seeks to understand contextual parlance, and thus accurately translate colloquial terms or dialectical slang. It is currently under beta testing and is at the final stage of ergonomic fine-tuning.

In relating other strengths of the device, May said: “It’s a fully independent translation earpiece. And what I mean by independent is that it doesn’t require any connectivity to your phone by Bluetooth or Wi-Fi. A lot of our competitors do.”

May was moved to design his own translation device after an incident during a business trip in China. His passport got stolen four hours after he arrived in the country, and when he tried to greet a police officer to ask for help regarding his stolen passport, a competitor’s translation service shocked the policeman when it said “Hello, I love you” instead of “Hello, how are you?” After four years of working on the tech, May now has his own language conversion device.

For his part, IBM Watson Master Inventor Neil Sahota relays the positive outcomes that they expect with the introduction of Translate One2One, saying: “It’s an absolute game-changer for the global translation market, offering significant potential for both commercial and social benefit, which we’ve only just begun to explore.”

“Lingmo’s latest breakthrough demonstrates the transformative potential that cognitive computing platforms like Watson can offer in solving some of the world’s largest and most persistent challenges,” he added.

Translate One2One needs IBM Watson’s artificial intelligence system, the Natural Language Understanding and Language Translator APIs. It has 3G mobile connectivity. For the translation process to commence, both parties need to be wearing the earpiece, otherwise no translation will occur.

Translate One2One will sell for $179 and is currently available for pre-order. An app version for iOS can also be purchased; it includes speech-to-text and text-to-speech capabilities for more languages.

Google and Skype have also initiated the development of translation tech platforms, but their efforts have yet to reach something as ambitious as an earpiece. (Related: Artificial Intelligence – Our Brave New Future?)

New York City-based tech startup Waverly Labs has designed a similar device to that of Translate One2One, which is called the Pilot Translating Earpiece, after getting around $5 million from its initial Indiegogo campaign last year. It is scheduled for market release later this October.

Sources include:




comments powered by Disqus