Why translating languages is about to become easier than ever
Google’s new updates to its translation app is the latest in efforts to make sci-fi translators a reality.
Whether it’s the “universal translator” of Star Trek fame or the weird but effective Babel fish from Hitchhiker’s Guide to the Galaxy, devices (and animals) for overcoming the language barrier have long been loved in science fiction.
Today, the reality might be closer than ever.
Over the weekend, Google announced that it will soon release updates to its mobile translation app, enabling the program to automatically recognize when a user is speaking in a popular language and turn it into text in real time, The New York Times reports. This is a leap forward from the app’s current version, which is limited to translating written text between 80 languages and hearing translations in a few well-known ones.
Google’s updates are the latest in a series of efforts by both technology giants and startups to make translating text and conversations as easy as possible. Just last month, Microsoft launched a preview of Skype Translator, “the most recent and visible example” of more than a decade of research in speech recognition, machine learning technology and automatic translation. The app currently allows for automatic voice-to-voice translation between English and Spanish speakers, with the goal of expanding to include “as many languages as possible on as many platforms as possible.”
Word Lens, an app developed by Quest Launch in 2010 and since bought by Google, has the ability to translate languages using only a smartphone: Users just hold up their phone’s camera to a sign and the app immediately translates the words. The app was adapted for Google Glass in November 2013.
Word Lens is useful for European languages, as it currently translates between English and German, Italian, French, Portuguese, Spanish, and Russian. Users who need help with Asian languages might benefit more from Waygo, which employs a similar camera translation process for Japanese, Chinese, and Korean.
Of course, these translating programs and devices have their limits. Google Translate still tends to garble long text translations, a snag that might have something to do with how the app is designed: Instead of having an algorithm designed to extract meaning from particular words and phrases, the program scours the internet for similar texts with which to pair the searched expression. The vast corpus, or database of language, that the app can look through includes everything from UN documents in six different languages to all the papers ever put out by the European Union since 1957, among many others.
Skype, too, struggles to work with a variety of accents and cadences, while Word Lens has trouble deciphering stylized fonts and handwriting.
Still, as The New York Times’ Quentin Hardy puts it, “those complaints are churlish compared with what also seemed like a fundamental miracle: Within minutes, I was used to the process and talking freely with a Colombian man about his wife, children and life in Medellín (or ‘Made A,’ as Skype first heard it, but it later got it correctly). The single biggest thing that separates us — our language — had started to disappear.”
And it sure beats having a fish squirming in your ear.