[Accessibility] Re: Can you help write a free version of HTK?

[grasch]

Hi!

First of all: If you really want the HTK maintainers to change their
license, I don't think writing angry mail is the best strategy.

To put this mail into context: I am the maintainer of the simon speech
recognition project.

Yes, the HTK license is an issue.

But take a deep breath and let's talk about this:
The HTK does what it is supposed to. It is stable, extremely well
documented and "just works".

When we were evaluating the existing software for simon it's not that we
didn't choose free alternative X because we were "tricked" by the HTK it's
just that if we had chosen something else, simon would most likely have
failed completely. If that means using the HTK to get started, so be it.


> [...] and HTK
> has already hooked Simon, making it impossible for Simon to fulfil its
> mission of helping people with typing
Please don't post misleading statements like this before you research the
topic a bit more thoroughly.

The HTK didn't "hook" simon;
simon 0.1 required the HTK for everything and this dependency was taken
into account at multiple places in the code - nothing too big but still.
simon 0.2 moved a lot of the HTK code in a central place and it could have
easily been replaced back then with something else.
Since simon 0.3 (the current trunk version) we even moved away from the
HTK file formats and use our own. Once it's time to compile the model we
ask a generic interface to do "something" with our generic input files.
Right now there is still only a HTK backend for this process but adding
another one should be fairly trivial.

Also, in simon 0.3 you can set up a complete voice recognition without
ever touching the HTK.

One reason for this is because the HTK license is far from being the worst
thing on the planet: It doesn't impose any limitations on the models
created with it.

You can find more details about my viewpoint on this in the comment
section of this article ("Re: Proprietary Software Should be Replaced"):
http://dot.kde.org/2009/08/22/simon-speech-activated-user-interface-kde


> Because of the HTK license, Simon is not going to be fully integrated
> into Vinux, or Ubuntu which is the upstream distro we test
> technologies for.  Simon built on HTK can never be included Debian, or
> Fedora.  In other words, Simon is dead, because of HTK.
Nobody told me that my project was dead...? The last commit was 2 hours
ago...


> Fortunately, we can freely read the HTK source code, and can learn how
> it works.  We can then go rewrite it, and hopefully do a better job.
Thats actually illegal. Nobody who did read the HTK source code is
technically allowed to work on GHMM because of copyright issues.

The law basically is that if you get your information from a copyright
protected source (that doesn't explicitly allow you to use it to create a
derived work) and create something similar then it boils down to the
question: Could you have created your derived work without ever looking at
the "original"? If the answer is no, than you violated the original works
copyright.

Disclaimer: I am not a lawyer but that's basically how I understood the
copyright law - at least in Austria.


> I
> propose we start an open-source effort to do exactly that, in order to
> enable Simon and other accessibility software to be freely used to help
> typing impaired people.  There is already a similar effort under way,
> with a proper license:
>
> http://sourceforge.net/projects/ghmm/
I have already contacted them a while back and they said that their
library was not geared towards speech processing and couldn't be used for
it neither now nor in the short term future IIRC.


> If anyone on this list is interested in leading a project to make a free
> alternative to HTK, possibly based on ghmm, please give it a try!
>  We can certainly help with testing, and maybe recruiting volunteers
> in Vinux.  What we can never do is include HTK in Vinux.

Well I think it's much more realistic to just use SPHINX instead of
Julius/HTK which works _right now_ and shouldn't be that hard to integrate
into the simon framework.

The only reasons its not there yet are:
1. Limited time on my part
2. Very little documentation available online

Regards,
Peter
(simon developer)