You’re speaking in one language and the person you’re conversing with is speaking in another language–but you understand each other perfectly. No, you’re not in the Tower of Babel, a science fiction movie, or an alternate reality. You’re in 2021. We are going to discuss about “In-Ear Language Translation” in tis article.
In-ear language translation has paved the way for real-time translation, interpreting diverse languages and making them accessible to the general public. With the use of AI, it’s possible to go seamlessly between two languages now in one conversation.
But does it really work?
Today, we’ll take a closer look at how real-time translation works, how in-ear translation is being implemented, and what the future holds for these devices.
What Are In-Ear Language Translation Devices?
In-ear translation devices are by definition AI wearables. But unlike smart ear AI “hearables,” their main purpose is not to track heart rate or blood pressure, or other health elements. Rather, they provide a different service–translation.
AI is the power behind in-ear translation devices. Older models would convert speech into text, then translate it into a different language. With the advancement of technology, AI now makes it possible to translate a word, identify that word in the source language, and then understand what’s being said, in order to convert it into another language. Once it identifies words, it can then identify phrases and then sentences, and so forth.
It’s a much faster method due to deep learning and neural networks–with the machine being able to learn languages through specific words. It’s becoming much better at deducting the context in which those words are used. It also has access to a large database of common words and phrases in various languages–the original input that the machine was fed. Thus, it should be able to convert supposedly seamlessly.
This real-time In-ear language translation technology has about an 80-95% accuracy rate, according to various sources. The WT2 Language Translator wireless earbuds, which we’ll talk about later, have a 95% accuracy rate. Some vary in reaction rate to translate, from 2-5 seconds, while some rely on the cloud, which delays the speech to translation reaction time.
How Does In-Ear Language Translation Devices Work In Practice?
We’ve talked about real-time translation, but how does it work with wireless earbuds?
Waverly Labs was the first to pioneer in-ear real-time translation devices in a conference about speech translation and wearable technology in 2014. Waverly Labs then launched in 2016 with the Pilot earbuds, the first smart earbuds with live translation. They sold 35k units and grossed $8MM in sales, according to the Waverly Labs website.
In practice, the Pilot earbuds, like the WT2 and the Google Pixel Buds, which we’ll take a look at, all rely on the cloud for translation. That said, the accuracy with which it hears speech and the turnaround time for which it translates and relays that speech into another language, relies on your internet speed. For its technology though, it’s still fast. But if we want to take a look at the differences between these devices, we must take an in-depth look at them.
So here we go:
Waverly Labs Pilot Buds: Language Translation
A startup that’s crowdfunded via an Indigogo campaign, Waverly Labs pioneered the Bluetooth earbuds called Pilot Buds. They’re the world’s first smart earpiece that specializes in real-time translation.
The smart device uses speech recognition to understand people speaking in different languages, at the same time the machine translates in real-time using AI translation. The buds are currently worth $179.
You download the app through your smartphone and the earbuds will work through Bluetooth technology.
It goes between a converse mode for speech recognition and listen mode for interpreting, and supports 15 languages and 42 dialects. It provides on-screen transcripts of the translations on your phone/computer screen for easy reading, which is downloadable too.
It comes with one earpiece for interpreting, and a second earpiece for wireless streaming of music, media, or phone calls.
Google Pixel Buds: In-Ear Language Translation
Google isn’t far behind with their cutting-edge Pixel Buds. They are, after all, the makers of Google Translate and Google Assistant, which are all integrated with Pixel Buds. The earbuds rely on the Google Translate app as its translation technology. It was the major exclusivity of these earbuds prior, but as of 2021, now all Google-Assistant-optimized headphones and Android phones can use this feature for real-time translation. All the same, having the same compatibility with the Google suite makes for seamless integration.
Like the Pilot, you have to pair them with Bluetooth technology to a smartphone. To listen and translate in real-time, you have to hold down one earbud. For Google Assistant help, you send commands by tapping the right earpiece. To converse, you can use the Google Assistant on the Pixel Buds, as well as the Google Translate app on the phone to automatically translate. The Pixel Buds can translate conversations in real-time in 40 languages, but Google translate holds 108 languages.
With the advantage of integration with Google Translate, Pixel Buds can go far with real-time translation. It costs the same as the Waverly Pilot at $179.
The Role of Translation Services In the Future of AI Wearable Translation Technology
Waverly Lab’s Pilot Buds and Google’s Pixel Buds are only two out of many in-ear wearable translation technologies in the market at this moment. There are also Time Kettle’s WT2 Plus earbuds, V Best Life Wireless Translating device, Pieko EarBud translators, and more.
But do they really work? At this moment, even these high-tech wearable devices have not yet perfected AI translation. Although they’re helpful for casual translation for business or leisure trips, in-ear wearables still have a long way to go to simulate real conversations between speakers of different languages.
Firstly, they only support two-way communication; second, the substantial lag time these devices take in translation is still cumbersome; third, they aren’t all that accurate in translating.
Although it’s becoming more advanced in a language context, AI translation in general still doesn’t take into account idioms or slang, or differences in grammatical structure from one language to another. And, as much as In-Ear Language Translation devices can change the playing field when it comes to interpreting different languages, there’s still the option of time and tested human translation.
Closing Note: In-Ear Language Translation
It is general knowledge that a power combo of machine and human translation still provides the highest accuracy. Translation services, which showcase the convergence of real-time interpreters with AI machine learning, have provided human and machine solutions for decades. As neural machines get better at translating, in-ear wearable technology will get better as well. But for now, human translation is still ahead of the game.