Raspberry Pi Becomes a Universal Translator

hola-me-nombre-david-conroy

We’re still about 150 years away from the invention of the universal translator by [Lt Cdr Sato] of the Enterprise NX-01, but [Dave] has something that’s almost as good: a speech recognition, translation, and text to speech setup for the Raspberry Pi that theoretically allows anyone to speak in sixty different languages.

After setting up all the Linux audio cruft, [Dave] digs in and starts on converting the guttural vocalizations of a meat speaker into something Google’s speech to text service can understand. From there, it’s off to Google again, this time converting text in one language into the writings of another.

[Dave]‘s end result is a shell script that works reasonably well for something that won’t be invented for another 150 years. The video below shows the script successfully translating English to spanish, but it should work equally well with other languages such as dutch and latin, as well as less popular language such as esperanto and french.

The season three story arc was an allegory for 9/11 and the lead-up to the invasion of Iraq, people. It was genius.

Comments

  1. TehMeh says:

    Did this before, but my app also used google voice to translate/send and recieve/translate SMS messages. Fun to text prank text your friends with foriegn message SMSs :P

  2. ChalkBored says:

    But does it do Klingon?

  3. Julien N says:

    aǔ vi povas lerni Esperanton ! Se vi ŝatas programi, vi ŝtatos tiun lingvon ! http://lernu.net kaj http://eo.wikipedia.org/

  4. Georg says:

    Google translate app for android is almost like the Enterprise gadget.

  5. t-bone says:

    as well as less popular languages such as esperanto and french

    Ouch!

    • Michael Chen says:

      I’m pretty sure they mixed up latin and french in the description

      …equally well with other languages such as dutch and latin, as well as less popular language such as esperanto and french.

      • Whatnot says:

        French is popular, and a hell of a lot more than dutch, esperanto only has one or maybe two people on the planet who still cling to it. Latin.. that’s only priests and students.

        • MeehhT says:

          Now hold on. English has an unhealthy amount of Latin words. Honestly things would make more sense if instead of using borrowed Latin, Greek, French, German, etc.., words we just used their direct translations. So Latin translators are best used to figure out what root words mean.

  6. Tom the Brat says:

    “You don’t want to know.” — Hoshi Sato

  7. DDevine says:

    I think the headline is wrong. This really is “Google Translate is a Universal Translator” – as all the heavy lifting is done by Google.

    Disturbingly – everything you want to translate is read (and tracked, and perhaps spied upon by a third party) by Google. So really, you have text that is reading you.

  8. defaultex says:

    Hmm. Mix this with one of those APIs that can analyze voices to produce a similar synthetic voice and you have something pretty close to a UT. The only problem however was the APIs I’ve seen that do that require lots of voice samples, the likes of which most people would lose patience and settle for text.

    • Greenaum says:

      Siri has human speech down quite well. I’m sure Google have speech synth, and for all I know it’s on my phone somewhere, but not well known. But since Google already does voice search, and good voice synth has been done for years now, it’s only really the equivalent of a couple of shell scripts at Google, or some API mixing, and they could hook up a Universal Translator in an hour or two.

      They really should. It’d be very useful, and get massive media attention, especially with the amusing Star Trek angle. Maybe wait til they’ve got something new to sell, or some bad news to bury.

      Actually is there a phone app for this?

  9. garym53 says:

    Ok, I have obviously missed something but what is the context/meaning of the line “The season three story arc was an allegory for 9/11 and the lead-up to the invasion of Iraq, people. It was genius.” underneath the video link?

    • Start Trek: Enterprise. Seriously underrated.

      Also, I’ve never read any commentary/criticism that makes the link between the season three arc and the US circa 2001-2003.

      • static says:

        However, while it was extremely under detailed, we just read such a commentary. :) Hell I had to look up what was meant by “season three arc” Dr. Who hasn’t been a TV program that I ever had an opportunity to view.

    • ChalkBored says:

      Pretty sure it was a reference to Star Trek: Enterprise.

    • garym53 says:

      Thanks all – since the show wasn’t specifically mentioned I didn’t make the connection – I suppose anyone who has watched the show would know “[Lt Cdr Sato] of the Enterprise NX-01″ was from that show – I also think there was probably supposed to be an asterisk linking the two references. I must admit I stopped watching all those addition Star Trek franchise shows due to the ridiculous “aliens” they had in them – the original series at least had the excuse of low budget, low technology, etc.

      • garym53 says:

        PS: most of what I know about the latter Star Trek shows comes from “The Physics of Star Trek” by Lawrence M. Krauss.

      • redfive1976 says:

        Except that Hoshi was an Ensign, not a Lt. Cmdr.

      • Greenaum says:

        It’s not so much the aliens as the “humans” that were unconvincing. Romance, power and tension, escaping death by the skin of their teeth (or more usually by reversing the polarity of some chronoton particles). All portrayed by people who seemed like they’d barely met each other. Memo: getting the guy with the crippling Asperger’s to write the scripts is probably not gonna work for a show you want to be a hit.

        And this is why Star Trek isn’t worth watching after TNG. Which was brilliant, had some great Scifi writers on board, and an excellent talented cast.

  10. static says:

    I have to believe French as a less popular language was written tongue in cheek. In a more perfect world Esperanto would be the go to second language learned. Although here in the US we would be slow to adopt it, hell we wont fully adopt to metric in any foreseeable future. Even though Google is doing the heavy work as another observed, this does make the job more lightweight.

    • Greenaum says:

      English is doing well as the world’s second language. Which makes it really easy on people born speaking it. I wonder if people whose second language is English learn it like we do French in school, or is it something they absorb from birth from all the popular media already in English? Do they grow up semi-bilingual? It’s probably Hollywood and TV that have spread English way beyond anything the Esperanto guy could do.

      From the few people who know, apparently Esperanto is a mess created by a guy who obsessively stuck to certain rules at the expense of easiness or consistency, and misunderstood some of the languages he was basing it off. Since then others have tried to make a better job of it, but English has the advantage of the massive existing base of material, as well as being the key to jobs working with the rich West. The only material in Esperanto is created by enthusiasts who want to spread the language.

      Still, as McDonalds, globalisation, and the bulldozing of global culture is showing, perhaps having everyone speak the same language isn’t the wonderful Utopia it was thought to be.

  11. Hack Police says:

    Arrest this guy, this is not a hack!
    ~

  12. HowardC says:

    It’s fantastic work, but the real problem comes from the fact that these api’s require a online connection. If google were only to release some of their online widgets in downloadable form we’d be good to go.

    • Greenaum says:

      And release their supercomputers and buildings full of hard drives onto a phone with an SD card. Tho they probably would work on a phone or PC, but then we’d have no reason to visit them so much. I suppose for the money, tho it doesn’t seem to be the way of the 21st Century to sell software for money.

      • MrC says:

        I think he means they already HAVE released this as a downloadable app.. I used it recently in china, no internet, but I could talk to people via the app, sure not perfect, and there is still the problem of when you really want to use it there is too much background noise for it to understand you…

  13. dx says:

    If something like this will work in real-time with device like google glass, so the device can translate on the fly what other people talking to user and translate response, it will be really cool.

    • notdave says:

      Minus glass, this is almost exactly what the Google Translate app does.

      You select the languages of the two people trying to communicate.
      One person starts speaking and the translated text is displayed upside down (so user #2 can see it).
      User #2 responds in their native language, and text is displayed to you in yours.

      • dx says:

        Yes, exactly, it almost ready for implementing in live real-time communication for example in foreign tourists trips (where internet is available) or in international conferences and meetings and so on. One more small tech step and what amazing effect! Of course it will work not completely perfectly, but it will be very usefull, if all translating process will be implemented in “voice form”.

        • Greenaum says:

          Perhaps it’s time that international roaming got easier and cheaper. It seems to be the way of the future. Still, until then, get your local friend to turn on their phone’s wifi hotspot, and connect your phone through theirs. Technology’s brilliant innit!?

          • dx says:

            Yes, of course. This can be used right now, and it’s already great hack. But using device like glass will make this feature more easy-to-use, more convinient and more effective (hands free). So it will be really cool and very usefull in hughe amount of applications, related to real-time translating.

          • dx says:

            The device like glass will provide almost transparent and invisible technical layer, so this feature will became almost your native, this is what I mean :)

      • dx says:

        And in addition to this glass can recognize the captions and texts in foreign languages in surrounding and translate them automatically – this is will be even more easy than voice recognition.

      • dx says:

        And, of course, not only Google can provide this service and equipment.

      • MrC says:

        what is really needed is a way to have a bluetooth/wired headset for you and the speakers/mic on the phone for the other person, so one side is English, the other German etc…

        Or another good idea would be a voice chat app that translated on the fly… workable for both long distance and in-the-room conversations…

  14. SkinnyV says:

    Less popular? French? 220 million of us are either offended, disagreeing with you or a combination of both right now.

    • MeehhT says:

      Google says it is only 74 million, which sounds right. Then again, it says there are 935 million that speak Mandarin and I read a long article on Chinese dialects that disagrees. Apparently it is more like 90 million, but because most Chinese family languages are not tracked outside China, everyone gets lumped in the Mandarin heading.

      Anyway, the point was to offend french speakers. Nothing is more fun than making French people irate enough to spew their sexy swear words a la (geddit) Matrix.

      • Greenaum says:

        The population of France and Tunisia, the two Francophone countries I happen to have been on holiday to, is 76 million. Add on Quebec, a big chunk of Africa, and wherever else. 7 billion people were in the world last count.

  15. Lol, your line about french has seemed to tick off a lot of uber-sensitive froggies…

  16. kommune78 says:

    People are still speaking french? Why?

  17. Eth says:

    Why being so harsh with esperanto native speakers? :-) :-)

    (disclaimer: I’m french)

  18. Rollyn01 says:

    Could it be that the part about French was a reference to the show Futurama?

  19. jazz says:

    there is hardware already for UNO models using(memes? phonems?) to get the input audio. You would need a synth(another arduino?) for the output. Now about that google…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 92,020 other followers