A Translator for All Languages – Does One Exist?
We’re an eclectic group of writers here at Nanalyze, and a fair number of us enjoy traveling around the planet while writing about technology investing. With some of our writers based in Asia, we’re also familiar with how much fun it is to live in a country where you don’t speak the language, you can’t read the writing, and at best you might find someone who speaks mediocre Engrish. In order to travel with no language barriers, we could just stop being lazy bastrds and learn the local language or we could try to find a translator for all languages. This might look like the following:
- A device that can translate in real-time what people are saying to us (ideally an earpiece)
- A device that translates our spoken words for another person to hear (ideally a smartphone emitting our spoken words in the desired language)
Unlike text on a web page which is easy to translate given the amount of localization most websites have, spoken language does not come with a drop down list of languages we can translate into. Whatever solution we use would ideally handle all languages to make things easy.
Real-time translation technology simply means that the technology allows a voice in one language to be translated into a voice in a second language within seconds after talking. The technology could come as an application or a combination of application and device. The size of the language technology industry is estimated at $30.96 billion, so consequently there are a number of startups that want to place translation technology in the hands of the man on the street. Let’s see how close we are to a translator for all languages.
A Translator for All Languages
Voice to Voice Language Translation
Experienced travelers would probably agree that the most appealing translation device would be a discreet earpiece that lets you understand whatever language is being spoken around you in real-time. Techies will recognize this device as the fictional Babelfish creature in “The Hitchhiker’s Guide to the Galaxy“. We talked before about a startup called Doppler Labs which is working on an earpiece that will eventually be able to do real-time translation. In the meantime, there’s been some buzz and a fair amount of skepticism around a New York startup called Waverley Labs that plans to offer the “the world’s first translating earpiece”. This startup has convinced a lot of people to pre-order $4.5 million worth of their $249 “Pilot” earbuds through a very successful Indiegogo crowdfunding campaign:
Pilot comes only in English, Portuguese, French, Italian, and Spanish or what we call the “romance languages” along with noise canceling features similar to the Doppler Labs’ Here One. Pilot is not yet available in the market but you can pre-order from their website. We’ve warned people about how easy it is to run crowdfunding scams so don’t be too shocked if something goes pear shaped here. If you want to wait until the device is released, maybe we can find a smartphone app that can perform the same functions.
This search led us to Sonico Mobile, an Austrian startup founded by Alexander Marktl in 2009 and the developer of a translation technology called iTranslate Voice. At the moment iTranslate is capable of translating 90 Earth languages of which 40 languages can be translated using real-time verbal translation. Over 60 Million people have downloaded iTranslate in almost every country in the world, and it’s being actively used by over 5 Million people every month. The application works online (or offline with the paid version) and includes features such as transliteration, conjugations, dictionary, and web localization:
In the nice heart warming video demo, we see a single female traveler approach a sweet Japanese lady and they communicate using the phone. Everyone giggles at how great the technology is and goes on their merry way. Let’s see how well that works when you stick your expensive first world smartphone in someone’s face who realizes that your smartphone is worth their entire yearly salary. What you might need then is a standalone device from a company called Logbar.
Founded in 2013, San Francisco startup Logbar took in an undisclosed round of funding in 2015 to introduce a translator for English-Japanese and English-Chinese. The translator dubbed “ili” is claimed to be “’the world’s first wearable translator for travelers” and is a self-contained device that doesn’t require Bluetooth or a Wi-Fi connection. Just like the previous translators, you just need to speak your language and the device translates via voice within seconds. Just watch this horribly annoying demo video to see for yourself:
Logbar intentionally started with just a few languages in order to start with a strong foundation. As we said before, if you can master the nuances of Engrish, you should be able to do just about anything. Just be forewarned that Logbar raised $1 million on Kickstarter before and delivered what some dubbed as the most disasterous product ever made. As we said before, you’re better off staying away from crowdfunding.
When you have translation needs while traveling, you’re not looking to discuss the merits of Einstein’s theory of relativity. You simply need to ask for directions, or find out what that food you just ate was. This means that the translation capabilities for a Babelfish type travel device don’t need to be that great. They just need to be easy to use and not likely to attract unwanted attention – like when you whip out your $600 iPhone and shove it in someone’s face.
Ideally we should be able to communicate using our voices over any platform and have real-time translation while carrying on more sophisticated conversations. Skype sees this as an opportunity and has developed their “Skype Translator” functionality which uses machine language to provide “near real-time” translation for 9 languages. That functionality is now available via the cloud so that other businesses can integrate it into their applications. Other companies like Apptek are also working on “automatic speech recognition” that supports multiple languages and is available across any platform you want to implement it on.
When we talk about a “translator for all languages”, we need to differentiate between one for simple conversions and one for more complex conversations. The former is already here, while the latter will still take some time to develop, even with the powers of deep learning.
Sign Language to Voice Translation
Lastly, we came across a pretty cool startup worth mentioning which has created a translator that does not deal with spoken language, but rather with sign language for the hearing or speech impaired. Although sign language is a bit different from the spoken language, it does have some similarities. It comes in as many variations as the countries and communities who use it. Plus, there are 70 million deaf people using sign language all over the world.
A company established in 2013 in Rochester, New York, called MotionSavvy, is trying to serve the sign language translation market with its tablet developed with gesture recognition technology. With about $500K in funding so far, MotionSavvy is working on a mobile device they call UNI which is a two-way communication that “reads” the sign language and translates it to spoken English. For the large population of deaf people, this can be a transformative technology. There are no dates yet set for the release of UNI to the market.
The key takeaway here is that if some of the world’s most prominent Chinese AI startups are still only able to convey their mission to us in Engrish, we have a very long way to go before we we can have a translator for all languages in text, much less voice-to-voice. In a previous article, we tried out the world’s best voice transcription software with less than impressive results. First, let’s nail the simple task of capturing voice to text, then let’s work on translating it in real-time for voice-to-voice. One step at a time.
If you enjoyed this article, then sign up for our free newsletter - Nanalyze Weekly. About every week, we'll send you a simple summary of all our new articles. If you didn't enjoy this article, share it on Twitter and tell everyone how much you hated it.