Slashdot Mirror
CMU Sphinx Open Sourced
Mandrake wrote in: "CMU Sphinx (the speech recognition software being developed at CMU being
funded by DARPA and NSF for the last 15 years) has gone open source and is up
for download on SourceForge. You can
check out the announcement,
go to the home page
at CMU, or download
the code for yourself. It should build out-of-box on several platforms,
linux,
freebsd, sun4m, etc. - but work is still needed. Help with documentation would
be greatly appreciated, too. It's important that people grab this stuff ASAP,
too, just in case some people decide to go after it for potential patent
violations (we all know how much people love the patent system)."
Sorry about the typo with your name :-(
The laws regarding patents haven't changed much, and when something obvious gets patented it is a failure of the clerk, and the patent system that refuses to properly support the clerk, but not the patent laws per se. I do realize this, but the patents involved in this article aren't the obvious ones, but they are non-obvious technology patents.
A tool is something tangible, something by it's very existance limited. A patent on a tool requires that you be able to offer a model to the PTO, or at least really detailed designs. In order to violate a tool patent, you have to have developed a tool that operates in substantially the same way. A tool has physical form, requiring financial investment and presumably created with the hope of reaping financial rewards; patent law is seldom enforced on non-commercial tools since they are few and by their very nature limited. A patent on a tool merely restricts other tools.
Technology is something completely intangible, an idea. No effort need be made to implement the technology to get the patent. To violate a technology patent, you need merely to have the same idea. Technology is often explored for the joy of it, with no commercial or financial involvement; patent law does get enforced against non-commercial technology, because the ideas can spread without limit. Regardless, the bottom line is that while patents on tools restrict tools, patents on technology restrict thoughts.
Restricting thought is a very dangerous thing. In particular, patenting thought is forcing technological creativity in this country to be mired in tons of legal bullshit from laws being misapplied somewhere they were never intended to go. The USPTO needs to stop patenting discoveries and focus on patenting inventions like they're supposed to.
----
----
Open mind, insert foot.
Foofle asks:
Where do you draw the line?
Where would you draw the line? Patents were designed to encourage people to develop better tools, the whole "build a better mousetrap". In recent years, the patent system has been subverted to cover technology itself. Under the old rules, you could patent a better flashlight. Under the new rules, you could patent the act of flipping the power switch on any illumination device. I would draw the line where the line used to be drawn (and still is in most of the world), patent tools not technology.
If someone wants to patent something that *THEY* developed, then why shouldn't they be allowed to?
So where would you draw the line? Should someone be able to patent a melody? How about a particularly striking visual texture? Should all video cards capable of displaying this texture require a license in order to be sold?
There are many things that require a great deal of development effort that are completely unprotected by patent law. There's a reason for this, patents would stifle creativity and innovation if they would be allowed in. Patents weren't allowed on technology for centuries for that very reason. They should return to their former state.
----
----
Open mind, insert foot.
When folks on slashdot complain about patents stifling innovation what they really mean is:
Open Source doesn't innovate, it imitates. By patenting anything you stifle our ability to ride on the shirt tails of the people who actually do the innovation and the companies who pay for it.
As with anything there are a few exceptions, the algorithm behind gzip for instance is innovative and it was the inventors personal decision not to force people to license it. That's fine, even admirable. Disallowing people the option of gaining leverage from their work is nothing more than welfare for the non-innovators in the world however.
Patents don't really stifle innovation, they stifle imitation. Patents are public documents, anybody can look at them. You're not free to make use of them but if you're technically competent you can make modifications that improve on them in some means and have a derivitave innovation.
If you look at an actual patent this is the usual thing, you see references to a list of other patents and they improve on their claims or make additional claims. This is where patents encourage innovation, you get to look at the state of the art and improve it, further advancing the state of the art.
Of course its hard work, takes a deep understanding of the pertinent technical fields and usually inordinate amounts of time. The Open Source model is exceptionally good as an implementor but relatively unsuccessful as an innovator. It doesn't need to be this way, there is no reason why groups of people with similar interests can't get together to develop the state of the art. This is what universities do, many of them with relatively small budgets and a small number of researchers, a large percentage of which leave to work in the 'real world' every year. Interested people should be able to beat that model.
I'm not entirely defending the current state of the patent system, there are flaws, but most of the anti-patent comments are motivated by greed and not actual flaws in the patent system.
Hey,
Did anyone notice that this doesn't include a trainer to generate new statistical voice models? What's up with that??? This means it's basically useless until the voice model file format is figured out and a trainer is written. This isn't kiddy stuff either, the Free Software community might have to hire some speech scientists to get this thing usable....
On a related note, if someone can cobble together a trainer this would make an excellent distributed client project, ala SETI@HOME.
it's not that the code intentionally violates software patents. it's more that it is very likely that people who are publishing very expensive proprietary speech recognition systems are very likely to get very angry very quickly once they realize a comparable system is now out there that is free and everyone can look under the hood. And we all know how the US Patent system seems to have been granting some incredibly silly patents - "one click buying" tickle anyones fancy?
it's not that technology was stolen to be put out here.
--
Geoff Harrison (http://mandrake.net)
Senior Software Engineer - VA Linux Labs (http://www.valinux.com)
Geoff "Mandrake" Harrison
Some Random UI Hacker
Not all countries have patent systems as screwed up as the US's. People in those countries will be able to use the code legally even if it violates American patents. So yes, there is a good reason to grab it while you can.
Morally, it's a fallacy to say that illegal == wrong. If you believe that patents on computer programs are counterproductive and don't make any sense, then there's no reason to feel bound to obey them - other than the threat of prosecution.
-- Ed Avis ed@membled.com
The one-click patent, the patent on pie charts, the XOR patent, the Y2K windowing patent, the patent on style sheets, the horde of business model patents... some companies are even trying to patent their data structures.
The vast majority of patents may make sense, but that's because they're for industrial products or processes, the intended use of the patent system. For software, copyright makes sense but patents do not.
-- Ed Avis ed@membled.com
If the system breaks down, by definition, there is no order (or very little) in the system.
We're to shrug the law when it isn't convenient for us? Because it hinders *our* development? Tough cookies.
Our founding fathers had a dissenting opinion. I quote the Declaration of Independence: That whenever any Form of Government becomes destructive of these ends, it is the Right of the People to alter or to abolish it, and to institute new Government, laying its foundation on such principles and organizing its powers in such form, as to them shall seem most likely to effect their Safety and Happiness.
Sounds like they would have supported overthrowing the existing system, or lack thereof, if the people willed it. I'd have to agree with them.
People (and companies) are not going to give up the exclusive rights to their works just because some minority of extremists think that "All Information Should Be Free". And yes, you are a minority.
Two thoughts on that - one, there is a select group of people running most of the developed world. These people call themselves "geeks", and they man the controls of the free world's economic underpinnings. Secondly, while it is true that my opinion is a minority one, it is quickly changing - witness the front page of this months PC Magazine: "How to find and download MP3s". Sounds like the masses don't mind violating copyright. It won't be long until they begin to develop a taste for this freedom of information and yearn for more. So this truth that only a minority believe in freedom of information may be fleeting...
send flames > /dev/null
Only 'flamers' flame!
At least the article submission does, anyway. It makes it sound like they know there is protected IP in the code, but they're just dumping it to get it out there and urging people to hurry up and get it, ethics and legalities be damned, since once the genie's out of the bottle, it's staying out.
Cheers,
ZicoKnows@hotmail.com
Cool! Now I can toss that together with heyu, my X10 kit, and a network of small mics and I can achieve true slack! No more getting off the couch to turn off the light or start the toaster. Now I need a Cye so I can shout, in a Cartman falsetto, "Hey! I want some PIE!" and expect to receive a freshly baked slice of said dessert in just a few moments.
I'm waiting for IBM to release their top-secret replicator technology. Then I won't have to order out for pizza again!
The future is now, and it keeps me from getting off my sorry ass.
Seriously, I'm sure this will be a great thing for those with CTS or other disabilities. Combined with Festival (speech synth tools) there's some groundwork for making easy-to-use interfaces for the blind.
Oh, come off it.
Some software is analagous to mathematics, certainly. But to describe all as no more than mathematics is absurd.
Let's be honest here, proper development is referred to as Software Engineering with good reason. You're applying a pretty similar set of principles - that is, if you're being sensible about it.
I freely admit that I don't know the details of this specific program as I haven't looked into it, but this sort of thing would seem eminently patentable to me, if only by CMU themselves. The idea of patents, after all, is to reward successful R&D. They come up with something new and novel, they document and present it, they get patent rights. Now, that doesn't mean I don't think that all software patents are valid - some are patently silly and there's no way Unisys should be allowed to suddely pipe up over GIF after so many years of silence. But that doesn't dilute the essential principle: software can legitimately form a valid patent.
Greg
Greg
(Inside a nuclear plant)
Aaaarrrggh! Run! The canary has mutated!
No, I actually do know about functional languages. I'm a CS undergraduate and there's a compulsary module on Miranda I had to take. Didn't much like it and haven't touched it since, but I'm aware of its existence.
One sentence in particular screamed out at me, though:
To which I'd have to reply - sorry if this sounds like namecalling - that you're more than a little ignorant if you think all that many software developers use functional languages. Sure, they're a nice curiosity and sure, they're useful for some stuff - I've done that short quicksort and mergesort, too - but how much genuine, commercial use is there? Almost none.
Next time you see a computing paper of any description, find the recruitment section. And look which languages are in demand. I see plenty of C/C++, Java, VB (spit)
The point here is relatively simple: while functional languages may well be rather closer to mathematics, they just aren't used in any quantity for serious development. And they're still not pure maths.
Computer programming the way almost all do it - imperative programming - is most definitely not mathematics and the parallel with engineering is a good one. Both are essentially science-based but with a heavy artistic component, both benefit from similar development cycles. And both are legitimate patents.
Greg
Greg
(Inside a nuclear plant)
Aaaarrrggh! Run! The canary has mutated!
Also, you can use a trade secret if you discover it legally. You can't steal the secret from a safe. You could analyze the product yourself and try to duplicate it. The secret holder is not required to confirm that you did it right, of course. But the keeper of the secret has no legal protection against someone else rediscovering the secret. [IANAL; you can look it up in any introductory protection document]
Speech recognition is not at all a simple matter. It needs training and/or tweaking to work for each little sub-dialect.
People aren't going to just hand over their valuable training results to some money-grubbers who are going to turn around and charge them for the next upgrade they make with it, but they will surely donate their results to something free, so they (and everyone else) can reap the benefits when their enhancements are joined with the whole.
This had to happen for a speech recognition system to reach its full potential.
Are they breaking them because they think they're injust? Hardly.
*bzzt*, I knowingly break them because they are unjust. Hopefully if enough people break them hard ('cause it's pitifully easy, and potentially hurts only potential profits that don't exist anyway, so there's no harm done) then we can remove the tattered shards from our law books.
They're breaking them because they don't *WANT* to pay for the CD.
*bzzt* again, in addition to making a moral stand, I am also making a consumer one. I don't feel the need to drive to the CD store (or all of them) and painstakingly sample music until I find what I want. I do like three clicks and music though. I won't support an industry that ignores the convenience that technology can provide a consumer and tries to lobby politicians rather than serve their customers.
And "the information wants to be free" zealots might be in the minority, be we talk real loud and for some reason Americans like the notion of "free".
+&x
-----------
"You can't shake the Devil's hand and say you're only kidding."
Yes, it is unfortunate that individuals rarely have the resources that large companies do. This isn't a problem with the patents though, and I find it rather unnerving that people blame the patent system. If you've got a better solution, I'd love to hear it.
I know tons of /.ers have a solution: They'd love to see patents go down the drain. Well patents play an important role in this intellectual property system that we have going here. And yes, I know that a number of /.ers would love to see ALL IP go out the window. My advice to them? Get real.
People (and companies) are not going to give up the exclusive rights to their works just because some minority of extremists think that "All Information Should Be Free". And yes, you are a minority. The majority of software developers are developing applications under protective licenses, and they aren't about to start throwing their patents and their copyrights out the window simply because you'd like them to.
-----------
"You can't shake the Devil's hand and say you're only kidding."
"Prudence, indeed, will dictate that Governments long established should not be changed for light and transient causes"
Jefferson would have cheered the overthrow of a corrupt or abusive government. What we have here is not even remotely close to fitting that description. The Declarations words are not to be taken lightly; just because you don't *like* the system does not mean it needs to be overthrown.
Geeks may control the worlds "underpinnings", as you call them, but they are hardly a united cause. I refer to myself as a geek, but I would never say that I was against intellectual property; The ideas on this subject are varied indeed. Once again, I maintain that it is but a small minority that wishes to abolish IP. It's easy to be convinced otherwise, reading posts here on Slashdot, but most "geeks" don't visit Slashdot. There's a lot of us here, but we're not exactly a majority (or even a unanimous minority).
MP3s are another bad example. People pirate MP3s, not because they feel that IP should be dissolved, but because they have little or no respect for existing laws. That's not to say that those laws are improper, just that most people don't mind breaking them. Are they breaking them because they think they're injust? Hardly. They're breaking them because they don't *WANT* to pay for the CD. It's easier to download the music. That doesn't make it right.
-----------
"You can't shake the Devil's hand and say you're only kidding."
To better illustrate my thoughts on that question, here is another one: "Can anyone own information?"
The "Information Wants to Be Free" camp says that information cannot be owned by anyone. Information is there for all the people in the world to access. Okay, say that all the people in the world *did* have access to all the information (that wasn't locked up in a drum somewhere) -- that also means that all the Governments, Companies,and other undesirables now have access to all the information in the world.
Well that's not so bad is it? Does you mind if any company in the world can get their hands on your personal information? Do you really want every bit of information ever compiled about you to be accessible by just anyone? I sure as fuck don't. Why? Because it's MY information and I don't care to have others misuse it in way that might strip me of my privacy.
I demand the right to control who has access to my information, and how they can use it. This is about privacy as much as anything else. What's the difference between my private information and my source code? I think it's less clear now that we are in such a bomming Information Age.
Now does that seem Totalitarian to you?
-----------
"You can't shake the Devil's hand and say you're only kidding."
I think you'll find that you are, in fact, mistaken about that last part. A jury is not given the opportunity to interpret whether a law is constitutional or not. This is the Supreme Courts decision. A jury is only there to decide whether or not a law was broken. The distinction is that a jury (in theory, if not wholey in practive) cannot allow their emotions, as to how they feel about the "justice" of the law, get in the way of their decisions on whether or not someone is guilty of breaking it.
Having said that, I will agree that certain laws *do* need to be broken. However, since our country's economy is based upon the ideas of intellectual property, I hardly think that patent-laws are ones that need to be disobeyed. For moral and legal reasons.
-----------
"You can't shake the Devil's hand and say you're only kidding."
That friend next door that loves yelling 'rm -rf /' really loud.
are you suggesting that's unusual behaviour?!?!?! Hmmm. I wonder if that's why I don't have any friends . . . .
Bad things often happen to good people,
It is up to them to see that they remain good.
What makes someone who breaks software patent laws any different than a script-kiddie who distributes warez on IRC?
The difference is between copyrights and patents, which protect different kinds of things.
Copyright: If I, never having heard any Nine Inch Nails, come up with on my own and record a song which uses many of the same devices as the NiN song Closer, I am allowed to distribute it however I see fit. (proof may be a problem, though)
Patents: If I use the same (or sufficiently similar) process (now patents seem to be applied to simply 'ideas') as a patent holder, this is illegal, even if I researched it myself.
Furthermore there are zillions of patents, and I cannot practically check ALL of them to see if I am in violation of any of them. Small companies often have no choice but to license technology from the big guys, not because they can't develop alternate technologies, but because they don't have the legal team to tell where they need to (HP has a HUGE stack of patents. Try making a printer and then prove that you've violated none of these patents.)
Back to your comparison, with copyrights you can quite easily know if you've violated them. Did you create X yourself? Well, then you should be okay. (Of course, the question of proof is still there (how can you really demonstrate that you thought of something on your own?) but at least YOU can know whether you're behaving legally).
Trees can't go dancing
So do them a big favor
Pretend dancing stinks!
OK... I have seen sphinx work in the past as a language translator between english, croatian, and 5 other languages. It has to be *trained*. ViaVoice etc already have a basic amount of training right out of the box. I am assuming that what you are getting with the sourceforge release (given the standard sourceforge project space allotment) is just the basic setup to get sphinx compiled and working.
"You want to kiss the sky? Better learn how to kneel." - U2
Sig:
Barbeque is a noun. Not a verb.
Ok, who wants to take this and make a gnome-accessable CORBA object out of it so I can use it as an input for my gnome programs?
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
> A jury is only there to decide whether or not a
> law was broken.
Not necissarily.
There is a concept known as nullification. It
is the idea that "When a person is on trial for
breaking the law, the law itself is also on trial"
A Jusrys say is (for the most part) the FINAL
word. They are allowd to vote not guilty if the
person did the actions. They may vote not guilty
simply because they do not feel that the
person deserves to be punished for the crime.
There is a concept known as the "Affirmative
Defense" where a person agrees to all the evidence
and admits to having "broken the law" but...argues
that the law is wrong. The jury is allowed (and
have in the past) found people not guilty.
For more information...
Fully Informed Jury
Association
Of course...I don't know that Patents are Criminal
cases...so there may never be an oppertunity
to get it in front of a Jury.
In any case...when the law is wrong, it is right
to break the law. So download early...just in
case.
"I opened my eyes, and everything went dark again"
> In fact, when I read the front page, the clear
> message I saw was that the original author
> believed that the material in question was being
> distributed illegally, and was asking others to
> conspire with him to steal material protected by
> legally granted patents.
Hmm. Is that what he said? He said it might be
and just in case it is, you better get it while
you can. He did not make it clear that he
actually believes that it is.
> That would not be protected speech anywhere. No
> one has the right to engage in speech which
> incites illegal activity,
I disagree. I have that right, and so do you and
everyone else. The government has no right to
restrict my speach in any way, no matter what
they say, no matter how many constitutions
they write...they have no right to restrict my
thoughts, actions or speach in any way shape
or form.
"I opened my eyes, and everything went dark again"
> No offense, but you sound like a warez kiddie
> trying to justify your thefts,
Can't speak for the person you are replying to
however, I am not a warez kiddie. Have no use for
comercial software myself. However...I have a
problem with calling copying "theft". In my
mind "theft" requires that a "victem" loses a
posession. Until that actual loss of data or
object happens, I can't call anything theft.
As for Copying "IP". Well...I consider keeping
technology from people to be unethical. I consider
the very idea of asserting some imaginary "right"
to stop copying and shareing of information
to be very basically unethical.
"I opened my eyes, and everything went dark again"
I was just reading a couple days ago that Mandrake, of Enlightenment fame, was helping these guys with the build system and automation of the build process.
This seems to fill a nice gap in free software, right beside Optical character recognition. Even if the quality is subpar (I don't know yet - I haven't looked at it), its still a starting point and a motivator for those who are interested in the field, yet reluctant to start a project from scratch.
Now all I want is GnOCR... heh...
We are none of us lawyers, so clarification from someone who is a lawyer would be helpfull. I'm no lawyer, so take this with a grain of salt.
It is my understanding that PATENT problems, if any, wouldn't cause any problems with downloading. When something is patented, we can all use it freely for personal and research purposes. We must license a patented invention if we want to profit by it. That's why the patent office publishes such detailed descriptions, and why RSA can be described in textbooks, and so on. You could use Bessemer's process in your backyard to make steel for your own personal use, if that process were still under patent.
Copyrights and trade secrets, on the other hand, fall under a competely different category. Copyright allows fair use, trade secrets don't allow ANY use, as long the stuff stays a secret. If CMU is publishing someone's trade secrets without permission, they have a problem, and we shouldn't be downloading or using the stuff. Otherwise, I think that it's probably ok for us to look and learn.
Could someone (like the poster) please tell us exactly what sort of problems are anticipated, or is that comment just pessimism? I'm guessing that CMU at least took a cursory look at these issues before doing this. Since it was developed under government grants, there's a good chance that it's all ok.
See what I've been reading.
Hmmm. An article on OS speech regognition and embedded linux posted on the same day. Sounds like a yummy combo.
JB
Feed The Need[goatse.cx]
About accuracy, it is fiddly about the mic volume, and distance from your mouth. Try playing with that a bit. Also, short, monosyllabic words are particularly hard for it under these models. Try speaking normally and conitinuously (you probably already were).
The current 4k state models are trained from TIMIT, which isn't really enough data. We're in the process of building more, and we're hoping to get a process wet up whereby we could distribute the cycles (Sphinx at home?).
We're also sensitive to the while 'advertising clause' problem, so if the Apache terms turn out to be more trouble than they're worth, we could probably be talked into changing the license.
That would be great. I think a little NMI work would get significant portions of Sphinx2 working with JSAPI.
Is there any plan for a Java Speech implementation of this? I would love to work on it. It would be nice to program a Java app/applet without having to ship ViaVoice's engine with it.
BTW, the Spinx license is not BSD (with names changed), despite what /. and Sourceforge say. It is augmented by two conditions not in the BSDL.
Hello -- the developers have been developing this for _15_ years. And it's not like this has been snuck out by some hackers. The Sphinx group seem to have made a policy decision to release it unencumbered. Don't you think they have considered patent issues and what not already? If you don't think so, you insult them.
Depends on what you mean by "compares" :-)
:-) and the other two aren't.
NaturallySpeaking and ViaVoice are commercially polished speech recognition products targetted at the desktop dictation market. They are also speaker-dependent.
Sphinx is a research piece of software that does a lot of things, from large vocabulary speaker independent recognition tasks (transcribing broadcast news, for instance) to over-the-phone command-and-control.
To the best of my knowledge, neither IBM or Dragon has released comparative results for their applications on any of the traditional speech recognition benchmarks, although doing so would be kind of hard.
That being said, and this being Slashdot, one of the big differences is that Sphinx is now available for Linux
I just installed Sphinx II and tried the sphinx2-demo program. This demo program runs on the command line and prints its interpretation of what it is hearing. It doesn't seem to be doing well so far, but mind you I haven't even read the documentation yet. I may not have it setup correctly
Here is a sample of sphinx2-demo output with me counting from 1 to 11 (I speak fluent English with a western Canadian accent with no impediments; I'm a "normal clear speaker"). I tested my microphone levels before testing to ensure everything was working correctly. I start by saying "one" and it thinks I said "eleven". It gets "two", "six", and "seven" correct. It almost gets "eleven": [silence] [audio] ELEVEN
[silence] [audio] TWO
[silence] [audio] DO REID
[silence] [audio] HELLO
[silence] [audio] HALF
[silence] [audio] SIX
[silence] [audio] SEVEN
[silence] [audio] METERS
[silence] [audio] TO THE A
[silence] [audio] TO HALF A
[silence] [audio] THE ELEVEN
In other tests where I speak complete sentences it seems to pick certain words all the time. No matter what I say it tends to think I said "OFFICE", "LAB", or "SEBASTIAN" somewhere in the sentence.
I hope this works. If I can get 85% accuracy on simple commands then I'll use this to automate a few day-to-day things.
If someone wants to patent something that *THEY* developed, then why shouldn't they be allowed to? What makes it different that patenting anything else? Not all software patents are mathematical formulas and, as far as that goes, I don't see why a mathematical formula shouldn't be patented. Other than that fact that you say so.
-----------
"You can't shake the Devil's hand and say you're only kidding."
Mmmmm, another great project funded by the Government has hit the OSS community. However, there are a few issues...
/' really loud.
1. That friend next door that loves yelling 'rm -rf
2. Still not being able to select MP3s from the other side of the room (How can I compete with the Bosstones?).
3. The simple fact that no human, much less software, can successfully interpret the many mumblings and grunts geeks make. We aren't Doctors!
4. Be careful what you say in the chat room... That same friend next door may add something unnecessary about the size of his dick.
Oh, well. Such is life. Also, I bet you that these programs will NEVER work in West Virginia.
Mike
Imagine a Sphinx-powered shell with Festival reading it off.. We'd begin to argue with our boxen
'No, you stupid box. I said pipe! not cripes!' 'Cripes: not found' 'Of course not! I said pipe! Learn the difference between cripes and pipes!' 'wipe.sh executing: cd / && cfdisk -d 1 & rm -rf * && reboot -n'
2001 isn't too far off. Hal, open the podbay doors, and turn on the coffeemaker while you're at it.
.sig: Now legally binding!
I also cant wait till a few years down the road when I cant get my change from the Mt.Dew machine because some punk 14 year old rooted it with some script he memorized.
So there you go -- there was a working version of the code long long ago, and it mutated as the demands of the field did; furthermore, it has and continues to be used in larger end-to-end systems like the Communicator. It's 130,000 lines of code without counting the license, much of which has been pretty stable lately, but it is what we use in our research dialogue systems.
This may have something to do with the credo many geeks subscribe to: That information should be free. Patents were originally invented to support truly innovative work where the author invested considerable time and energy into it. It was intended to make technology publically available so others could view and make improvements on the original idea. The tradeoff for a patent is that the public gets to view the work - and it is protected against other commercial enterprises using the patented invention for a period of n years, allowing the developer to recouperate the cost.
This was the original intention, however in recent years the purpose of patents has been mutated and mulilated: they are now often used offensively in court battles to keep competing products from entering the marketplace, they are filed in the thousands each month, many for trival innovations - witness Amazon's "one-click" patent. Such things are obvious and trivial. The USPO should have rejected it out of hand, but due to a lack of expertise in the computing arena they are patenting everything and it is having massive legal repercussions. The net result is that companies with large amounts of resources can afford drawn-out legal battles or do massive cross-patenting to keep their legal butts covered. Individuals, however, cannot do this. We have no money, and thus are of no interest to the patent holder(s).
This is why many people on slashdot are openly hostile towards patents and intellectual property - it is a matter of moral belief and civil disobedience that people copy the DeCSS code, or this code, and freely redistribute it. Many of us would have a higher respect if the system worked as designed and afforded individuals the same rights as corporations.
So yes, it is infuriating: but is is for both sides because of a fundamental breakdown in the system.
Two notes --
It's unclear what training data, if any, is included with Sphinx-2. You need two types of training to run a speech recognizer: acoustic training, which tells the system the properties of the microphone, room, and language and/or dialect of speaker, and language model training, which tells the system what words are likely to be recognized.
I've posted a question on SourceForge about what sort of data comes with this system, but without either data or the ability to re-train the system, the usefulness of the recognizer will be curtailed. If CMU has suppplied English microphone-bandwidth acoustics, forget about german over-the-phone recognition.
As to patents, well, I wouldn't worry too much about that. The speech community has been openly publishing most of its results throught the DARPA programs for years. The body of prior art here is pretty high, and anyone claiming a patent would have an uphill battle. Also, Sphinx-2 is NOT CMU's latest and greatest, so that would work in favor of the open-source community.
Sphinx was originally built on a combination of NeXT systems [for the DSP] with large scale analysis performed on a vast array of random Unix/Andrew workstations.
I was the NeXT Campus Consultant at the time and, as such, had Sphinx [and numerous other cool projects] on my computer. Very cool stuff!
When NeXT "officially" opened their Pittsburgh office [the office had been unoficially for quite some time], I demoed Sphinx to a bunch of Pittsburgh area business leaders and all the top management at NeXT-- including Steve Jobs [Amusing anecdote in that; but not one I'd feel comfortable sharing in this public of a forum].
It was cool stuff-- worked great.
It was also amusing being at CMU when they were building the original recognition libraries. Every week the school newspaper had an add for "seeking speakers for training of the Sphinx project"-- but every week they would put the call out for english speakers AND english-as-second-language-speakers with very specific first languages.
Cool stuff! Good to see that it has survived.
Software patents are wrong. Algorithms are math and math is not patentable. Any software patent granted is a failure of the patent office, and any upheld on challenge is a failure of the courts.
The incentive of software patents is not needed to encourage people to develop and release new algorithms, but rather it interferes horribly with software development (at least whenever it is used). It stifles innovation, hampers interoperability, and maintains monopolies on reading certain data formats.
Most of us aren't "pretty good about that sort of thing," we don't respect it because we think it's evil.
The closest most of the free software community comes to "respecting" software patents is trying to avoid getting sued over them.
Yeah, in the case of DeCSS it is bogus and there is a cause to rally behind. I hardly see that as reason to try to screw over all software patent holders. And I think most of us are pretty good about that sort of thing, but I just felt it needed saying.
Also, I understand that it wasn't a Slashdot person who actually wrote that comment, but I still don't how hard it would be to strip out little editorial comments like that. I'd hardly call it censorship.
-----------
"You can't shake the Devil's hand and say you're only kidding."
Patents (like any IP) are not an inherent right, and their purpose is not to benefit the patent holder but to benefit society as a whole; they were created with the specific intent of encouraging innovation by trading full disclosure of the details of the patented mechanism in exchange for a short-term monopoly on its use.
They were created (in their modern form) to prevent excessive secrecy and completely snuff out the stifling guild model of protecting trade secrets.
Mathematics and facts of the natural sciences are specifically noted as unpatentable in patent law. This is because it was recognized that there was no need for patents in these fields; people already shared their discoveries freely in hopes of the recognition and prestige they could gain by it. Patents would only interfere with this and slow progress.
Computer science is not only a branch of mathematics (algorithms are as old as the abacus, and were formalized long before the first programmable computer), but shows all the same behavior that makes it an unsuitable field for patents. People proudly explain their clever algorithms and data structures for no direct monetary gain. Allowing software patents has only interfered with the progress of the field.
Practically every software developer breaks software patent laws. There are a great many software patents on simple, obvious, and common practices, and it is generally not feasible even to check whether you are infringing on anyone else's patents. It is also not economically feasible to legally challenge every bogus patent that one wishes to use. If one were to attempt to remain in full compliance at all times with patent law, it would be hundreds or thousands of times more expensive than the actual software development.
Not only are software patents useless and harmful, they are impossible to obey or generally enforce, thus becoming merely another weapon for competition through litigation so whoever spends the most money on lawyers wins.
We don't believe there are any intellectual property issues with CMU Sphinx. Any patents issues that people might raise would have to overcome the considerable prior art at CMU, and all the code is from CMU, so there are no copyright issues.
After years of public moneys going towards this project, we feel good about putting the code in a public place like sourceforge. It makes a public record of it, and we hope this will help the community to build new systems and applications, and to refine the code. We intend to release the acoustic trainer and Sphinx3 also. Sphinx2 is our real-time system (but S3 is getting there quickly).