DARPA Project Babylon: Universal Translator

silance writes "Take a look at this project from DARPA (Defense Advanced Research Projects Agency)! This time the boys are trying to hammer out a portable, two-way, real-time, multi-lingual audible speech translator proposed to be run on everything from PDA's to wearable military hardware to workstations (to replace their PRE-EXISTING ONE-WAY real-time hand-held audible translators, of course!). The site contains descriptions of technical approaches, a technical milestones timeline, and a nifty Power Point presentation for the executive-types ;) They should give William Shatner a beta model out of pure respect... Here's a link to Google's cached HTML version of the Power Point presentation just in case. (P.S. - get a load of that logo at the bottom of the page!)"

Comments

pattern recognition?

[Telastyn]

The next big thing I think would be a "smart" translator that can do pattern recognition and "learn" as it gets more of the language. IIRC This is how the star trek translators work.

Kind of the difference between pattern checking, and anomaly detection in virus scanners.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Concerns...

[NanoGator]

It's cool that they're working on this and all, but their promises of building these into PDA's set off a flag in my mind. There's another company that, as of a couple of years ago, had developed a realtime program that allows one to speak english into a mic and have spoken japanese come out.

I remember reading that they needed serious processing power and RAM to make this work. (At least 512 megs...) It seems like if one language takes up this amount of resources, then it'll be a while before we have a multi-lingual PDA...

Maybe their technique is different? I dunno. I know it's not the same company.

I guess I'm just concerned about this being vaporware.

Nice

[Auckerman]

Initial impression: boy are they in a hurry. Very aggressive time table for this project. 6 Months to "Emergency DARPA", 18 Months to 3 functional prototypes.

Then I saw what languages it will have: Arabic , Mandarin (the part of china that border Pakistan and India is mainly Islamic), Pashto (Pakistan/Afganistan), Dari (Iran/Afgan/etc)

Oh. What I want to know is what those 8 other languages are that they want to have the ability to add to it later?

interesting logo, but bad.

[Artifex]

The "Babylonian" reference may at first seem apt: the towers were built 'to the heavens' (well, pretty high) and a lack of communication and understanding among peoples led to their downfall.

However, the underlying, unspoken subtext of a comparison between us and Babylon is that we displeased God. Remember, in the Bible at least (there's other versions in other histories/religions), God was displeased, and the language confusion among the peoples was caused in order to bring us down.

What this logo basically tells the world (or at least those who have an understanding of the mythos) isn't that we're a great nation and metter communication would have helped us - it's that we went against God, and this is how we paid.

This sounds a lot like those right-wing extremists who tried to blame the attack on 'communists' and homosexuals in our country making God upset.

Now, I feel, like many people do, that our country has done a great many things wrong: setting policy based on oil needs and not human rights, keeping some smaller countries' governments (including some democracies) destabilized in order to serve our own interests, etc. However, just as I don't think that we can claim "God is on our side," neither do I think anyone can claim that God isn't.

This logo is offensive. That it shows the half-thought-out mentality of some of the people in charge at our governmental agencies should be a cause for alarm, not applause. We have been called Babylon by many people with grievances against us, and it seems our leaders are reveling in the name.

Re:interesting logo, but bad.

[cosmicg]

I used to work on a NASA research project, and the building we worked in had a large sculpture of Icarus outside the backdoors. At first I thought it was a poor choice, but now I think it is well placed and powerful, serving as a constant reminder against technological hubris, or at least using wax in heat sensitive projects.

In the biblical myth of the tower of Babel, god introduced multiple languages to prevent people from working together. If you ignore the whole sinners-working-against-Yeowah angle, this DARPA project is basically allowing better communication and cooperation, overcoming the obstacle god enforced at Babel. Factor in the fact that much of the current round of fighting/terrorism is the result of religious convictions (or just "god"), and you've got a deep, thoughtful, and symbolic choice for title art.

Of course, somebody probably just thought it looked cool, but...

Re:Give it to Hoshi

[uebernewby]

True. Hoshi's fiddling with the universal translator really made me think about that piece of equipment we've been taking for granted in previous Star Treks.

Seems my university syntax and phonology courses weren't *that* useless after all...

The way I see it: suppose Chomsky's Universal Syntax turns out to be not innate to human brain structure, but to the very essence of communication. Meaning: if you're going to communicate something, all the forms you're going to be able to do it in will conform to a fairly basic set of ground rules and all the intricacies of natural languages are simply icing on the cake, as it were. If you figure out what that Universal Syntax is (sorry, I forgot the exact term he used - it's been a while, and my university education was in Dutch), you can feed that into a computer and teach it to reduce all phonemes from a given language to it. Then you can have the computer expand the basic message back into coherent communication in another language using the same basic rules.

It's late. And when it's late, this is the kind of stupid stuff I think about.

Oh, and I don't think Hoshi's *that* cute.

N-Gram

[rufusdufus]

For those who don't know, an N-gram is data structure which encodes the statistics of word order in a language. These are used to greatly improve the accuracy of language pattern matchers such as speech recognition.
A typical speech recognizer might use a 3 word N-gram (tri-gram), which keeps track of all probable words which follow and thier likelyhoods. The probabilities are calculated by running terabytes of english text (books, magazines, internet chat boards) through a word counting program.
Thus, "green eggs and" will get a very high probability for "ham", but low for "jam", so it can bias a sound that seems to match "mam" acoustically to the more likely linquistic match "ham".

Re:Babylon

[Idylwyld]

It wouldn't make sense to name these systems "DE" babel whatever. The system in no way reunifies the languages (assuming a biblical sundering of language from a unified root). In fact in all likelihood the success of a system such as this would result in even greater differences in language, especially if the system could learn. Imagine a translator that could render a foreign language understandable and maintain a slang similar to what you used. Self-reinforcing deviations from "proper" grammar could become a serious problem.

already been done for Russian <=> English

[js7a]

A St. Petersberg company called Ectaco has already done this for bidirectional English-Russian handheld speech-to-speech translation. They call their stuff "Universal Translators" too.

Re:mmm......

[AntiNorm]

or maybe haXor to newbie

Google already does haXor, so maybe this isn't so far off.