Slashdot Mirror
Copier Auto-Translates Japanese to English
StCredZero writes "Wild. Fuji has created a photocopier that automatically translates documents from Japanese to English. That's pretty nuts. Apparently, the copier can figure out what sections are text, OCR the text, send it to a translation engine, and put the english back into place."
Turn into actual pictures of people, too! Amazing!
They've been using this for years to translate instruction manuals.
Imagine if you upload manga scans to Flickr, and it automatically translates them to English.
Imagine if you upload anime to YouTube, and it automatically includes an English subtitle.
Virtual Betting on Facebook for non-geeks.
StCredZero "writes wildness. Fuji drew up the photocopying machine which automatically translates the document from English from Japanese. That is the clean nut. With respect to appearance, as for the copier the text, as for OCR what kind of section text, to send that to the translation engine, and in the place English". You reset, or can grasp.
main(c,r){for(r=32;r;) printf(++c>31?c=!r--,"\n":c<r?" ":~c&r?" `":" #");}
It translates to Engrish?
I reserve the right to think for myself. Others' opinions are optional. Puppy on lap = typos...not illiteracy.
It doesn't seem to be mentioned in TFA, but I have to wonder: Exactly how fast does it copy if it has to translate? I'm sure it's not the near-instantaneous work we've come to expect of our Xeroxes. If the translations aren't just gibberish Engrish, its usefulness will be immense, so the time won't be so much of a concern; but I do still wonder.
98% of America's teens drink alcohol, smoke, and have sex. Put this in your sig if you like bagels.
Between the inaccuracy of unproofed OCR and the poor quality of machine translation, I can't imagine that the results are very good.
LOAD "SIG",8,1
Hey, I just got one of these and all it'll print is "All your base are belong to us."....
Don't tell me to get a life. I'm a gamer; I have LOTS of lives!
[Reading from his book, "Jimmy James: Macho Business Donkey Wrestler," translated to Japanese and back again]
Jimmy: I had a small house of brokerage on Wall Street. Many days no business comes to my hut. Jimmy has fear? A thousand times no. I never doubted myself for a minute, for I knew that my monkey-strong bowels were girded with strength like the loins of a dragon ribboned with fat and the opulence of buffalo...
[pauses while turning page]
Jimmy: dung.
...for a machine, under the current paradigms (that is, no true artificial intelligence) to properly translate something. Translation is not an exact science, and you can't expect to get a decent translation by just having a word-per-word approach. Heck, not even a sentence-per-sentence or paragraph-per-paragraph approach would ever be enough. Translation requires deep social knowledge--you need to know what you are translating, from whom you are translating, for whom you are translating... that is, you need to enclose your translation in a sociological context. No machine can ever wish to do that without artificial intelligence. It's hard enough as it is to get a human being to understand that word-per-word translation is stupid--imagine telling that to your CPU core.
;)
Disclaimer: I'm a translation student myself
They probably used it to translate the instruction manual into Engrish.
Most of the stuff on
This may help people from making some big mistakes with their tats...
They did something like this with a Japanese car manual a long time ago; it ended up something like: "If a passenger of foot should obstacle your passage, tootle the horn. Tootle him melodiously at first, but if he continues obstacles your passage, then tootle him with vigor."
...before we see photos of the scanned documents in http://www.engrish.com/ ... What were they thinking of??
The Wknd Sessions - Malaysian and South East Asia independent music
Our those which achieve the main thing of the ether of title in order to know never clearly that remnant because of the remainder which it should find!
Since I could never have created the above err, prose, myself, I typed the following answer into babelfish and translated it into Japanese, and for good measure, back into English.
Clearly that remains for those of us who have achieved the title Ether Lord to know, and for the rest never to find out!
And as you tread the halls of sanity, You feel so glad to be, Unable to go beyond. I have a message, From another time..
Either you are old, or a bit naive. I think in the next 10 years we will see significant improvement.
Yeah, 'cause researchers have long promised us that AI will reach us in 10 years. <sarcasm>
Seriously, I think you underestimate the difficulty of translating. Have you done any major foreign-language translation -- especially of conversational speech? My experience has primarily been with Japanese and English, and I'll tell you right now that it can be nightmarish.
Sentence fragments are the worst part. Japanese has a completely different word order from English. All modifiers (including phrases and clauses) come before the word they modify, and the language has a Subject-Object-Verb order. "I just saw the man who stole my friend's watch last Tuesday" becomes "Just I Last Tuesday friend's watch stole man saw." Now try translating that from Japanese to English when the sentence is cut in half.
Worse, the language has very different levels of allowed vagueness. "Complete" sentences in Japanese can contain just a descriptor or an action without any specification of who did/was what. Conversely, translating "3 of them" in English to Japanese is hard because you have to know "3 of what?" to know what counting suffix to use.
Another problem is that many very different words sound exactly the same when conjugated to the gerund or perfective forms. English has a number of homonyms, but there are MANY more opportunities for mix-ups if you don't have access to kanji to tell the semantic meaning apart because Japanese has a much more limited range of phonemes. For example, take "katte" which is the gerund form of the verbs "kau" (buy), "kau" (keep/raise), "karu" (cut), "karu" (spur on), and "katsu" (win). That's 5 completely different verbs that conjugate to the same sound. If they're written phonetically or your going from speech, then you have to be able to understand the meaning behind the words to translate. (Did I mention earlier that you may not have an explicit subject and object to go off of?)
Then you get into issues of translating things like politeness levels, different ways of addressing people, and other concepts that don't translate well into English or concepts like singular vs. plural that are dropped in going to Japanese. Let's not even consider puns and poetry!
These are not trivial issues. An automatic translator would need to somehow be able to conceptualize what a person is trying to speak about, which would require understanding the story being told and an ability to predict where they are going with it. This will require strong AI.
Accurate and intelligible translation is an art -- not a science -- because it requires an intuitive and empathetic ability to understand the mind of the speaker well enough to map their thoughts into a different method of expression.
If it's for-profit but free, you're not the customer -- you're the product (e.g., the Slashdot Beta's "audience").
What is wild is that anyone with half a working braincell would use a photocopier in an office where a copy of every document is sent to an uncontrolled 3rd party for translation.
Yeah, put that baby in the CEO's office..
(not the mention the fact that there's a huge gap between mechanical translations and the subtleties of language only a skilled translator and/or native speaker has any hope of translating).
So, IMHO cute idea, but don't expect me to bu one any time soon.
Insert
Biblefish translations to spanish says something like This building _is being looked at_ by Surveillance Cameras (lit.) Any illegal activity will be registered and taken _the_ police. So, you actually register the activities, but take the police away, not the activities to the police.
Better is the portuguese translation: This building is taked attention by Surveillance Cameras. All illegal activity (that is allright for brasilian portuguese) will be recorded and the police will be examined.
Italian and german seem to be more or less OK. The french translation is also understandable.
Francisco
Obviously the translator was all at sea.
Pining for the fjords
That's the fun part of these automatic translators - put in perfectly good English, translate to a foreign language, and translate back to English. Good fun for a rainy afternoon.
"Today is under construction... please do not be alarmed by the construction men hanging themselves from outside your balcony. We will take them down tomorrow..."
Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
These guys saying that the technology won't be here within their lifetime have to be ancient or just forgetting how rapidly the pace at which technology accelerates has been increasing of late. How long ago was it that this here "Internet" only had a few hundred nodes?
I am not exaggerating when I say that automatic translation from extremely dissimilar languages requires strong AI. You need to be able to guess what a person is thinking from what they're expressing to map it into a different way of expressing themselves. You also need strong AI to understand the flow of conversation when terms are not expressed strongly.
As an example, Japanese doesn't really have a word that maps to "it." They have a word that maps well to "thing," but nothing that matches "it." This is because pronouns in English fulfill the function of referring back to a concept expressed in a previous sentence to place it in short form in the context of the sentence being expressed.
English Example:
E1: Hey Frank, did you buy that TV yet?
E2: Yeah, I bought it yesterday.
"Japanese" Example:
J1: Hey Hiro, already you that TV did buy? [or still you that TV haven't bought?]
J2: Yeah, yesterday bought. [Note the lack of "I" and "it!"]
Languages like English make translation easy in this regard because you have a generic pronoun to "hold the place" of a specific subject or object. You don't have to know what it is -- you just fill it in. In Japanese, if the conversation were to continue about the TV, the word TV would never be brought up again until the subject of a sentence changed away to something else.
Often, you can have a conversation in Japanese where the subject is never explicitly spoken because it's obvious from the context of the speakers. Given the frequency of homophones in the language (particularly in conjugated forms of verbs), this can make translation maddening if you don't know what the speakers are talking about (because you can't see what they can see or don't know what they know).
This is most frustrating when you're dealing with an author who is using ambiguous or cryptic speech by off-screen characters to give a sense of foreboding or foreshadowing. The conversation is just as cryptic to a native Japanese speaker as it is to us, but we literally cannot translate it to English without knowing the secrets ahead of time because grammatically correct English cannot be that vague!
Anyway, I'm starting to veer off from my original point which is to say that accurate translation requires modeling of the minds of the speaker which requires strong AI. A simple dictionary + grammar rule-set or even a theoretically complete database of possible sentences and phrases will never be able to achieve translation because of the inherent differences in the levels of specificity in the two languages that requires you to model and understand the thoughts and intentions of the speakers.
Frankly I've mostly given up on strong AI within my lifetime after so many decades of empty promises, so I don't see accurate automated translators coming any time soon.
A final thought:
While I've harped on the difficulty of going from Japanese to English, there are some tricky parts of going the other way -- I just don't have as much experience. The one time I wrote a letter in Japanese for a class that included words I didn't know beforehand, I ended up accidentally using words that sounded bizarre and in one case insulting because words in different languages don't map 100% to each other. A word that means the same thing in Japanese and English for one use may not mean the same thing in another. For example, you can use both "karu" and "kiru" to mean "cut" when talking about hair, but you'd use "karu" for mowing grass or shearing sheep, and "kiru" for chopping up fish and accidentally cutting your finger. The relationship between words is a Venn diagram, and computer translation gets that wrong when it's unable to realize what the (omitted) subject or object of the sentence was.
If it's for-profit but free, you're not the customer -- you're the product (e.g., the Slashdot Beta's "audience").
Its powder level! It's over 9000!
(I tried so hard not to rise to the bait... I honestly did. Then I thought of the power/powder pun and I couldn't stop. Damn you Psykechan! Damn you!
That's the fun part of these automatic translators
Actually, this may not be automatic. Japanese and English are (I understand) two languages which differ on fundamental levels. It's not like trying to get to the station and using Franglais: "poor alley a la gar?" It's totally different. Translation is very, very hard, and words are used differently, too. My last job involved working from a spec written in Japanese, and translated by obviously intelligent Japanese people into English. It was... interesting... work, and involved asking the Tokyo office the same question multiple times in different ways, and then cross-referencing the answers, to see if they were consistent.
Also, one of my favourite Japanese poems reads like this:
"She said she would come
At once, and so I waited
Till the moon rose
In the October dawn"
That's from "One Hundred Poems From the Japanese"; I have a different translation in another book, which comes complete with an explanation of over 100 words (I kid you not) of all the different interpretations of the Japanese original, and why that specific translation was selected. It turns out that the original Japanese is written to be gender-ambiguous, and person-ambiguous. It's not "she" and "I", but something like "unresolved consious being" waiting for "mysterious, gender-ambiguous lover-person about whom I care".
You can see why they settled for a snappier version for the poem translation.
So while I love the meloncholy of the translation, I also feel a bit cheated by the fact that my language won't support the original concept, which sounds even better (when you make up words for those concepts and then replay them in slow-mo in your head).
Also, it makes me very glad not to have to translate instruction manuals or business specs.
'No rational religion claims "supernatural" exists, that's an atheist slander.' - seen on slashdot.
Actually, an interesting point from what I wrote above: I spent about 2 months (on and off) trying to work out what the Tokyo office wanted. Many emails, and quite a few phone calls were made. Then we sent out a guy to finish off the code (he'd started it in London, based on what we THOUGHT was required). He had my research and English spec (and his preliminary code) as the basis of what was going to be done, but basically, *he* sorted the final stuff out on the spot, in less than a week. It turned out that a lot of the stuff I thought was wrong, or not *quite* right.
It emerged (and I *never* thought I'd say this, let alone write it) that having a coder on the spot, saying "if we use these rules, you'll get THIS result" turned out to be *more* efficient than getting the spec right beforehand. More accurate, as well.
I simply could *not* nail down an accurate spec for 2 months, because of the problems of translation. The Tokyo guys didn't speak perfect English. I speak no Japanese. We were all intelligent business people; I'm a geek, and a Japanese guy was a geek. We couldn't do it. The only language which worked, at the end of the day, was the logical programming language which spat out results, and the analysis thereof.
When we got to THAT level, the business people could finally say: "change this to Y", or "this value should be X". They had the Japanese technical spec, the system output, and the business knowledge. Only then could we resolve *every* issue.
'No rational religion claims "supernatural" exists, that's an atheist slander.' - seen on slashdot.