Japan Getting Real-Time Phone Call Translator App
another random user writes with news that NTT Docomo, Japan's largest wireless carrier, will be rolling out a real-time translation app for phone calls on November 1. At launch, the app will translate Japanese into English, Mandarin, and Korean, and later that month it will add French, German, Indonesian, Italian, Portuguese, Spanish, and Thai. No word on Klingon. From the article:
"The products have the potential to let companies avoid having to use specially trained multilingual staff, helping them cut costs. They could also aid tourism. However, the software involved cannot offer perfect translations, limiting its use in some situations. ... It provides users with voice translations of the other speaker's conversation after a slight pause, as well as providing a text readout. ... NTT Docomo will soon face competition from France's Alcatel-Lucent which is developing a rival product, WeTalk. It can handle Japanese and about a dozen other languages including English, French and Arabic. The service is designed to work over any landline telephone, meaning the company has had to find a way to do speech recognition using audio data sampled at a rate of 8kHz or 16kHz. Other products — which rely on data connections — have used higher 44kHz samples which are easier to process."
One step closer to Star Trek. Seriously, though, someone needs to set up a "BadPhoneCallTranslation.com" domain and set it to replay some of the funniest or awkward phone call translations.
And on an unrelated note, I am now seeing stories with red bars at the top. Am I now seeing articles to be posted in the future? Or is this just another Slashdot weird redesign?
See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
Ever notice that none of these stories are ever written from the jobs perspective? "I lost my translation job because ___ company is rolling out a software program that will do my job for them."
Repeat until there are no jobs left.
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
I'm still waiting for an english-only voice processing software package that can translate boston and texan accents, and god help you if you have a southern drawl, or english-as-a-second-language. Poor bastards who speak Russian can't 'V' to save their soul.
#fuckbeta #iamslashdot #dicemustdie
Yet another reason for the Japanese not to learn English properly. As if we didn't have enough Engrish already!
Hope there will be a Welsh translation service it will make working for the UK Gov. much easier as everything including help lines has to be translated into Welsh as well as English :(
Build a Man a Fire, and He'll Be Warm for a Day. Set a Man on Fire, and He'll Be Warm for the Rest of His Life.
Nope AC. Of course there's the gap.
The job goes away as fast as the slip in the Manager's office. Rent's due next week and you needed that paycheck. The "Opportunity" doesn't show up for years.
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
Sore de ganbattekudasai!
xkcd is not in the sudoers file. This incident will be reported.
Or maybe the translator gets fired and two unemployed salesmen get hired because with this new technology a company can start rapidly expanding its international markets. Don't pretend you can predict the future.
limiting its use in some situations.
Don't worry humans, most of you are not going to lose your translation job.
We already limit human translators to "important" stuff only.
The people most likely to lose their jobs are Indian call centers. Soon, you'll be "talking" to Bob at Dell who walks you thru power cycling your windows box, but he's actually in Afghanistan and doesn't speak a word of English. Also instead of telemarketing scum leaving messages on my answering machine, they'll be having Turing test conversations about how I should vote for any politician but Johnson (whom I am voting for).
I'm sure there will be a contractual limitation not to do anything important with the service "So I'm not telling ya all, that ya all can't not shut off the backup reactor cooling pump disabling relay..." WTF does that mean in English much less Japanese? So... no critical infrastructure support, no medical, no legal, no engineering, no management, no HR, no accounting... whats left other than telemarketing and call centers?
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
*call from hungary*
....
- Hello?
- This record is scratched.
- My hovercraft... is full of eels!
I was hoping it was powered by brainwaves, and involved cramming a fish into your ear.
It seems that the devaluation of language skills would be a corollary of costs being cut.
After all, if companies won't pay extra for language skills, why acquire them (while you're in high school)?
Ironically, these sorts of apps may lead to lower levels of foreign language ability within a given society. Instead of having millions of French, Chinese, and German speakers in, say, the US, you would end up concentrating language skills in a few computer systems (Google's, Apples, maybe Wolfram's).
I'm not a lawyer, but I play one on the Internet. Blog
Can you imagine how rich this dude would have been had he actually patented every concept he came up with for Star Trek? Fortunately, back then you couldn't patent a concept, because our government wasn't as corrupt.
And therein lies the ultimate irony of Star Trek; for everything Gene got right about the tech, he failed miserably predicting human nature and greed.
If telephones are outlawed, then only outlaws will have telephones.
I wonder what you'll have to say into the phone in japanese to get "all your base belong to us" out of it.
IF you translate between Japanese and English it would be FAR FAR FAR easier to do it with text first. Have you seen some of the translations google comes up with? Often it ain't even good enough to get the gist of the original message, let alone be good enough to carry a conversation.
With speech to speech translation, you got to first have proper voice regonization, a far from perfect technology, then have good translation, which doesn't yet exist especially for such totally different languages as japanese/english. And then you got have good text to speech. Which ALSO still isn't anything to write home about or computer game would use it to be able to forgo voice acting and have far more flexible scripts including usage of player chosen names.
Combine all three and engrish.com better get some extra hosting to store all the hilarious screwups.
The reason Japanese/English is so hard to translate is that the languages are completely different in which the way you say something and what information is included/excluded is totally different. And then people make it worse by leaving things out they expect someone from their own circle to know but that someone from another culture needs to make things clear. How would you translate 'dude'. And yes I know that is a lousy example of hip and happening street talk but I am a nerd, what do I know of being hip. But the translating engine better knows or you are going to get a tower of babel.
If this new solution magically improve all three fields needed to an as yet unheard of standard, it would be amazing. I doubt it, but it would be amazing.
Now just convince the japanese to stop putting all the text on their webpages in images and we might actually be able to make the world a bit smaller. Oh and get companies to make deal with global payment services so anyone can use their local payment system to pay anywhere in the world.
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
I currently use the best commercial voice recognition software, and have experience with a variety of machine translation suites that do Japanese-English, so I have somewhat of a frame of reference.
With respect to the former, the quality is quite good. I can speak at a relatively natural pace and get upwards of 95% accuracy. That said, I still have to adjust my speech in a sometimes unnatural manner to be sure the program "hears" me correctly, and I have heard horror stories from people with different accents/dialects having a terrible time. (Someone made a joke about Welsh, but I know a Welsh fellow and he had some colorful things to say about the commercially-available VR programs.) An additional complicating factor in the J-E scenario is that Japanese has many words that sound nearly identical, and are distinguished only by slight inflections: (hashi; chopsticks) and (hashi; bridge) is a typical example. There are thousands of examples of such words, and from what I understand, Japanese voice recognition software is quite far behind because of this particular trait. Without a UI for speakers to choose which word they are actually trying to say, I can imagine that the VR side of this program has a slew of problems.
Now, onto machine translation. As it stands, MT works great for some language pairs, but Japanese-English is notoriously problematic. AFAIK part of the reason is the highly contextual nature of the Japanese language. Subjects and objects are often omitted entirely, for example. I don't really have to go into this in detail -- just run any Japanese Wikipedia article through Google translate and see what happens. Other commercially-available and proprietary software I've used has been basically the same (Google actually seems to be a bit better, usually.) English-Japanese is a bit easier because the context (subjects, objects, verbs) are typically "all there" -- so even if the result is Japanese that is horribly unnatural, you may still be able to get the info you need.
But now, they're going to take the VR-generated input of varying degrees of accuracy, and run THAT through MT software that butchers even the most simple and perfect sentences? I could be wrong, but I'm having trouble seeing how the result will be anything less than a disaster. "The software involved cannot offer perfect translations, limiting its use in some situations" sounds like the understatement of the year. Get ready for synthesized-voice gobbeldygook and an mp3 website called "spoken Engrish."
(Full disclosure: I am a translator, but in a lot of ways a yearn for the day when MT will be good enough to put me out of a job; I think the idea of people being able to instantly communicate with speakers of other languages is exciting and would lead to a much more open world. However, I've been hearing that tech like this is 3-5 years away for decades now. If this product showcased a revolutionary engine for MT, then I would be singing a different tune, but for now it seems like a mere combination of two imperfect technologies.)
Voice recognition/transcription is again growing in popularity recently, what with Google and Siri. But it is still very lacking. My Google voice voicemail transcriptions are positively laughable and I don;t use Siri because I'm sick of arguing with that bitch!
Are we really expected to believe that systems that cannot reliably transcribe voice are now going to transcribe and translate on the fly? Forgive me if I doubt the effectiveness of this endeavor. But, think of the savings for the company. Instead of paying employees to actually respond to customer needs, they can now keep the customers busy arguing with an automated system, billing monthly and doing no work. I wish my business could abuse customers this way and still make me billions.
Tokyo, Japan, NTT DoCoMo, Inc. --- October 1, 2012 it is the world's first commercial mobile service for the translation of the conversation between people who speak other languages and Japanese on November 1 today announced that it will start, I express Hanashite Hon'yaku (called automatic) translation services. Also DoCoMo, today, by placing the camera on your smartphone in front of the text simply, he announced (translator of AR camera and word recognition) launched on October 11 of Utsushite Hon'yaku convert the signs and menus foreign.
Yeah, I won't hold my breath for this kind of technology until 10 years from now.
Tough luck. Nobody has an obligation to guarantee a job for you. You need to make sure that people will want to employ you - even as things change and technology advances.
All green of skin... 800 centuries ago, their bodily fluids include the birth of half-breeds. For the fundamental truth self-determination of the cosmos, for dark is the suede that mows like a harvest.
It's not just the lack of subjects/objects in Japanese that causes a problem with (machine) translations, it's also the different sentence format.
English, as well as many other European languages use the Subject-Verb-Object form, while Japanese uses the Subject-Object-Verb form.
I've also seen translators having trouble with negative sentences or modal forms (can/must/must not/can not), f.e: http://translationparty.com/#10538744 Japanese, as I've heard those share a similar grammar.
Odd... a part of my text disappeared:
However, humans don't have such a hard time with grammar, it can usually be picked up quite fast (in comparison to learning a whole new vocab and a "hard" character set). As far as these languages are concerned, it's all about vocab, and if it can pick it up and infer proper meaning from context, it could be useful. It may even work better for Korean Japanese, as I've heard those share a similar grammar.
One of the reasons why written J to E translation (sometimes, sort of) works is because the Kanji (chinese characters) can unambiguously dstinguish between homonyms. For example hana (flower) and hana (nose) sound exactly the same, so in spoken conversation you have to rely on context. But in written Japanese, they use different chinese characters, so a translation to English can be more accurate. There are a LOT of homonyms in Japanese, so many that it's one reason people have resisted abandoning Kanji altogether.
Given how bad written translations usually are--even with the Kanji to help, I can't see this as having a lot of accuracy.
The author originally wanted to use the New York Yankees as the focus... However, he was unable to capture enough examples of the Yankees making contact with a baseball during the ALCS to complete the study. ha
Reading about this made me want to reread this classic cyber-punk novel
Eric Carlson MIS Director Midstate Manufacturing
I want to say that your post was very enlightening to me (I'm a self-learner). Until now I did not know about the pitch accent in Japanese and thought that it was impossible to distinguish words written the same way in Kana if they were taken out of context. Getting the right word from the context doesn't seem to be this hard from the first glance, actually, since most words with the same writing have very different meanings and can be distinguished by analysing n-gram frequencies or using other similar techniques.
Yes yes, my nipples explode with delight too!
Pitch accents in Japanese won't usually tell you which homonym is the one that you want. Nor is it always easy to tell from an isolated sentence. Example:
Watashi no hana wa akai desu.
Either "My flower is red." OR "My nose is red."
Konojo ni hanawa wo agemashita.
Either "I gave my girlfriend a floral wreath (flower ring)" OR "I gave my girlfriend a nose ring".
The pronunciation of these is EXACTLY the same. Good luck figuring out which is correct without written kanji to help.
The Japanese forms for "must do" are very idiomatic and basically translate literally to "if you don't do this it's no good = you have to"
kore wo tabenakereba narimasen desu.
kore wo tabenakereba ikenai desu.
(Literally "if you don't eat that it's unacceptable = You must eat it.)
"Can do" is usually done through a verb conjugation with "e"
Nihongo wo hanashimasu I speak Japanese
Nihongo wo hanasemasu I am able to speak Japanese
Nihongo wo hanasu koto ga dekimasu I am able to speak Japanese (a bit more formal, literally "Speaking Japanese is possible")
Anyway, Japanese grammer is basically NOTHING like English.
Hah. Ask a Mandarin speaker how to say "Does the mother scold the horse?"
Every language has homonyms and near-homonyms. A Japanese friend who is a relatively fluent English speaker is continually tripped up by there/their/they're, especially when more than one form appears in the same sentence. Eventually machine translators will have to do a better job of what people do, which is decoding the the ambiguities based on context. Not just because different words can sound so similar, but also because most real, live, people don't enunciate clearly except in formal settings. You don't eat with a bridge or drive over chopsticks, and if you actually said something like that you'd most likely confuse native speakers. I'll consider machine translation "here" when it can properly translate a sentence slurred by a drunk person from Osaka without any configuration other than "Japanese".
I doubt it will produce anything useful for anything more complicated than, "Hello. How may I help you?"
Ever use Google translate for Japanese -> English or English -> Japanese ? Word salad is typically the result.
So clearly, the solution is to first have it translate the Japanese to Korean, then from there to Dutch, then German, and we can probably make it to English from there. Problem solved!
*Here's hoping this post doesn't get butchered in two*
By "can" I didn't mean being capable of doing something, but rather being allowed to do something, which I think is what's assumed by the modal forms.
So it would really be:
(You can/are allowed to speak Japanese)
nihongo wo hanashite mo ii
as well as other -te forms (for mustn't),
hanashite wa ikenai
(Musn't speak..)
and so on.
Also, this is wrong:
kore wo tabenakereba narimasen desu.
kore wo tabenakereba ikenai desu.
Should be:
kore wo tabenakereba narimasen.
kore wo tabenakereba (ikenain desu) or (ikemasen).
PS: Why doesn't slashdot.org still have unicode when slashdot.jp (clearly) does?
The sentence format shouldn't be a problem as long as both languages are internally consistent. Missing information is a much bigger deal. And Japanese is a worse than average language about leaving things to be implied by context.
Awesome in theory, but I wonder how accurate the service is. In order to trust it, I'd have to be fluent in the language myself. I'd hate for something to be misinterpreted, especially since the program isnt able to comprehend context. blog.telelanguage.com