Back in 1978, science fiction author Douglas Adams made his main character stick a fish into his ear in order to understand an alien language. The fictional (and now legendary) babel fish provoked plenty of laughs back in the day. Fast forward to the present, however, and people all over the world are sticking devices into their ears in order to understand foreign languages. Perhaps Mr Adams knew more than he was letting on.
Ever since reading The Hitchhiker's Guide to the Galaxy as a child, the concept of the Babel fish has stuck with me – no doubt in part due to my own fascination with languages and translation. It's something that springs to mind every time I sit down to write an article like this, looking at translation technology trends. Anyway, with that little piece of my childhood shared, I'm going to crack on with looking at some amazing translation technologies that you should look forward to using in the near future.
First, though, let's take a brief trip down memory lane.
At the core of every app, gadget and platform that we are looking at today, lies the concept of machine translation. It's a concept that dates back to the 1950s in practical terms, although theoretical discussions of machine translation date back as far as the 17th century.
It was in 1954 that the Georgetown experiment saw a machine successfully translate more than 60 sentences from Russian to English. One of the earliest recorded machine translation projects, it caused a flurry of excitement and set the expectation that machine translation would be available in just a few years. Funding poured in yet results didn’t follow.
In fact, it wasn't until the 1980s that any really significant progress was made. At that point, advancing computer power opened up the field of translator technology once more. Statistical machine translation once more fed the hope that we were close to solving the issue of producing successful automated translation. Unlike rule-based machine translation, statistical machine translation used analysis of bilingual texts as the basis of its approach.
Again, however, the result was not the natural-sounding language that human translators were able to deliver. Mistakes, poor grammar, clunky sentences and chunks of text that simply didn't make sense were commonplace.
It was the growing availability of deep learning applications and the use of artificial neural networks that really took things to the next level. Neural machine translation models entire sentences, using artificial intelligence to predict factors such as word sequencing, with far better results than previous models.
This has opened up major interest in automated translation once again, kickstarting a whole new generation of language translation technologies.
So, are technology translations now as good as those produced by human translators? Not quite yet, but they're getting close. When using technology translation still lacks some of the nuance and subtlety that human translation delivers. However, the fact that translation can now be undertaken so cheaply and at scale means that neural machine translation has taken off in a big way.
In 2017 for example, the European Patent Office began using neural machine translation technology to convert its documents into other languages. Since then, the system has translated more than nine million documents, with 31 languages available.
A wide range of translation apps and gadgets are available for both business and personal use, harnessing the power of neural machine translation. Let's take a look at some of the most popular.
Do you need to translate on the go? From business meetings to holidays, there’s plenty of use cases for in-ear language translation technology. Some of the most exciting in-ear products on the market at present include:
Like the fictional babel fish, the WT2 Language Translator slots neatly into the ear in order to provide real-time translation. The set comes with two earpieces, one for you and one for the person you want to talk to. The earpieces connect to your phone via Bluetooth, but you don't need to be online in order to use them, which can come in handy.
Users of the WT2 enjoy access to 40 languages and the nifty little device understands 93 different accents. Languages include Cantonese, Arabic and Greek. The accuracy is impressive, coming in at around 95%.
Mymanu CLIKS are another wireless earbud option that can operate offline. The brainchild of Ghanaian-British entrepreneur Danny Manu, the earbuds can translate 37 languages wirelessly when paired to a smartphone. The near-real-time translation (it’s a sentence or two behind) is of high quality and the ear buds are proving popular in the UK, the US and Asia.
The range of ultra-portable translation technologies that you can fit into your pocket is just as impressive as the variety of in-ear offerings on the market. A couple of popular options include:
If you’re after instant, offline translation, you’ll want to take a look at ili. It’s been designed for English-speaking travellers who need to communicate common travel words and phrases while on holiday. It translates them quickly and efficiently into Mandarin, Japanese or Spanish.
Unfortunately, ili is not without its limitations . The most significant of these is that, while ili translates pretty competently into each of the three languages I mentioned above, it doesn’t translate back into English. As such, if the person you’re trying to communicate with replies to you, ili won’t be any help when it comes to you understanding their response.
Hugely popular in Japan, the Pocketalk is a neat little handheld device that translates 74 languages from 133 countries. It performs well across a range of dialects for languages such as English, Spanish and Chinese, understanding not just formal language but also slang, idioms and even profanity.
Pocketalk can text your speech to the person you’re talking to or say it out loud, depending on your preference. The translation is fast and accurate, supporting a good pace of conversational flow. You’ll need a data plan to use the Pocketalk.
On a par with Pocketalk, Lincom provides real-time, two-way translation for 138 languages. Handily, it also includes a camera that allows you to instantly translate signs, menus and other written documents.
The Lincom shows translations on the screen (which is great from a language-learning perspective) as well saying them aloud. It needs to be online for access to all of its available languages, though 14 of those languages are also available offline, which is a nice touch.
Designed to be ultra-handy, the Lincom also allows you to record voice notes and even acts as a wireless hotspot, should you need one.
If you’ve got a bunch of documents to translate, there are some decent language translation technologies on the market to help you out. While human translation still has the edge, meaning you’ll want to partner with a translation company like ours for projects such as translating a marketing campaign or a sales letter, there are plenty of situations in which using a translation technology platform will suffice. You can read more about the blending of human and machine translation technologies by clicking the link below.
If you need high quality translation, but also need to cut costs, it’s worth noting that you can use a post-editing machine translation service to polish the results of machine translation. Doing so adds that all-important human touch.
Let’s take a look at a couple of established front-runners in the field of document translator technology.
Read more: Machines and humans working in harmony – is this the future of translation?
Blending statistical machine translation with rule-based translation, SYSTRAN allows businesses to translate documents rapidly and cost-effectively, meaning that it’s a popular choice for companies looking to cross linguistic barriers on a budget.
Another real-time document translation platform is Lionbridge, which covers everything from formal business language to the latest jargon. This can be a cost-effective option for companies that need to translate items such as blog posts and social media posts, though there’s always an element of risk when it comes to using machine translations on your social channels without getting a native speaker of the language to at least cast their eye over the translated content.
A stalwart of the language industry, SDL XMT is used by businesses around the world. The system is constantly learning and adapting based on user feedback, which means it goes from strength to strength. SDL XMT can handle both formal and informal documents and is available to translate documents between a wide range of language pairings.
Finally in this roundup of amazing modern translation technologies, I want to look at a couple of apps. I'll skip past the obvious choice of Google Translate, to look at a couple of serious contenders with interesting features.
SayHi is an app with plenty to offer. It's free to use, covers 90 languages and dialects and allows you to adjust the speed of the translation when you record your voice. Available for both Apple and Android devices, it's a great app to have to hand on holiday, as well as to use in professional settings.
If you're heading off on holiday to Asia and need an app that translates Chinese, Japanese, Vietnamese and Korean, then Papago has plenty to offer. It also translates Russian, German, Spanish and Italian, as well as English, making this a really versatile app.
As well as translating spoken language, Papago can work with written documents, including handwritten documents. It also features helpful image translation functionality.
If you're after cultural information, as well as advice on etiquette for visiting a particular country, then TripLingo is the app for you. You can download a pack for over 100 countries, which will enable TripLingo to translate the language for you and support you to learn it through a fun quiz mode.
The past decade has seen leaps and bounds in both translation technology and the ways in which that technology is made accessible to businesses and individual users. Apps, earbuds and pocket-size devices mean that we can take our translation tools with us wherever we go. Meanwhile, the range of online and offline options mean that even patchy internet coverage is no longer a major obstacle to making yourself understood overseas (and understanding those who speak back to you).
Over the coming years, we’re likely to see a whole range of other applications of this technology. There’s plenty of interest in combining automated lip-reading with translation, for example. There is also a lot of work underway around translation and regional accents. We’re already seeing the results of some of this work in the devices and services mentioned above, but the future should see even more translation apps and gadgets become capable of handling regional accents and dialects.
As the use of artificial neural networks and deep learning models continue to enhance the quality of machine translation, we can also look forward to the creation of new apps and gadgets to deliver that translation. That means more speech-to-speech and speech-to-text services on the market, as well as the traditional text-to-text platforms.
How many of these language translation technologies have you tried? Is there a device that I haven’t listed that you think is worthy of a recommendation? And what do you believe the future holds in store in terms of amazing translation technology? You can leave a comment below to share your thoughts, experiences and predictions. You can also click the link below to enjoy some recent musings on machine translation trends for 2022.
Post your Comment