Click here to register.

Languages

About VoxForge

VoxForge was set up to collect transcribed speech for use in Open Source Speech Recognition Engines ("SRE"s) such as such as ISIP, HTK, Julius and Sphinx.  We will categorize and make available all submitted audio files (also called a 'Speech Corpus") and Acoustic Models in GPL format. 

Why do we need GPL Transcribed Speech?

In order to recognize speech, Speech Recognition Engines require two types of files: the first, called an Acoustic Model, is created by taking a very large number of transcribed speech recordings (called a Speech Corpus) and 'compiling' them into statistical representations of the sounds that make up each word. The second is a Grammar or Language Model.  A Grammar is a relatively small file containing sets of predefined combinations of words. A Language Model is a much larger file containing the probabilities of certain sequences of words.

Problems with Current Approaches:

Acoustic Models are Closed-Source 

Most Acoustic Models used by 'Open Source' Speech Recognition engines are 'closed source'.  They do not give you access to the speech audio (the 'source') used to create the Acoustic Model.  If they do give you access, there are usually licensing restrictions on the distribution of the 'source' (i.e. you can only use it for personal or research purposes). 

The reason for this is because there is no free Speech Corpus in a form that can readily be used, or that is large enough, to create good quality Acoustic Models for Speech Recognition Engines.   Although there are a few instances of small FOSS speech corpora that could be used to create acoustic models, the vast majority of corpora (especially large corpora best suited to building good acoustic models) must be purchased under restrictive licenses.

As a result, Open Source projects that want to distribute their code freely must purchase restrictively licensed Speech Copora that limit distribution of the 'source' speech audio, but allow them to distribute any Acoustic Models they create.

VoxForge will address this problem by providing all Acoustic Models and their 'source' (i.e. transcribed speech audio) in GPL licensing format - which requires that the distribution of derivative works include access to the source used to create that work.

Restrictive Licensing Creates an Access Barrier to Potential Contributors

Every project that wants to build an acoustic model using a corpus with restrictive licensing must purchase their own copy.  This is difficult for FOSS projects, which usually have no revenue.  If a project does purchase such resources, the license restrictions will require them to keep the resources behind some kind of access barrier restricted to official project members.   This takes away freedom and flexibility from end users and shrinks the pool of potential contributors to the project. 

Acoustic Models are not Interchangeable 

Most Open Source Speech Recognition Engines ("SRE"s) come with an Acoustic Model.  However, these Acoustic Models are not interchangeable with other open source Speech Recognition engines.  The way to address this problem is to provide the 'source code' for the Acoustic Models (i.e. the Speech Corpora used to create the Acoustic Models), and permit users to 'compile' it into Acoustic Models that can be used with the Open Source SRE of their choice.  

VoxForge hopes to address this problem by creating a repository of 'source' speech audio and transcriptions, and by creating Acoustic Models for each of the main Open Source Speech Recognition Engines (such as Sphinx, Julius, HTK and ISIP) .

Open Source Acoustic Models Need to be Improved 

Current Acoustic Models used by Open Source Speech Recognition Engines are not at the level of quality of Commercial Speech Recognition Engines. 

VoxForge provides a central location that can collect GPL speech audio and transcriptions.  As more speech audio data is collected, better Acoustic Models can be created, to the point that someday they will be comparable to Commercial Speech Recognition.

No Open Source Dictation Software

Most Open Source SREs are designed for command and control and IVR telephony type applications (e.g. Sphinx, HTK and ISIP).  The Julius Speech Recognition Engine was designed for dictation applications, however the Julius distribution only includes Japanese Acoustic Models.  But since it uses Acoustic Models trained using the HTK toolkit, it can also use Acoustic Models trained in other languages - like English.  We just need hundreds of hours of transcribed speech audio to create English dictation Acoustic Models.  This same audio data might also be used to permit the other open source Speech Recognition Engines to work in dictation applications.

Although the current focus of VoxForge is on Speech Recognition for IVR telephony applications or Command and Control applications on the desktop, when the amount of audio data collected reaches a certain threshold, this data can then be used in the creation of Acoustic Models for Open Source Dictation Applications.

Why GPL?

Unrestricted Licenses for Speech Corpora will not be Effective

We believe that making Speech Corpora available using an unrestrictive, BSD style license will not help the Open Source Community in this particular case.  A BSD style license permits users to distribute derivative works without having to contribute the source of those modifications back to the community.  In our opinion, the Open Source Speech Recognition community does not have the required threshold of users to create a self-sustaining community using a BSD style license.  If there was a larger community, then there would be a greater likelihood that a self-sustaining group would give back to the community, even if not required to do so using a BSD style license.

GPL licensing ensures that any contributions made by the Open Source Community to VoxForge will benefit the community.  This is because the distribution of any derivative works based on the VoxForge Speech Corpora must make the source (i.e. the transcribed speech audio) available to the community.

Comments

AddSearch

By anonymous - 9/27/2013 - 1 Replies VoxForge is a very nice project. It would be cool if there was some sort of status page. It could

By AlAmir - 2/11/2012 - 1 Replies Hi sir,

By biswajit.net - 2/17/2011 - 1 Replies Hi,

By Arthur Chan - 1/21/2011 - 1 Replies Hi all,

By speechrecguru@yahoo.com - 7/5/2010 - 3 Replies i need an open source speech rec model for my business. I can contribute 100,000+ dictations and associated transcriptions to the project (medical reports with all patient and physician information redacted).

By shirish agarwal - 4/30/2010 - 1 Replies Hi there,

By kmaclean - 11/12/2009 In reponse to this post on LWN.net: Simon - speech activated user interface for KDE (KDE.News):

By softtalk - 10/8/2009 - 2 Replies How does the GPL license for voice data affect the owner's obligations with respect to application source code that uses the voices? In general, GPL licensed code "infects" any code that links to it. If my code requires voxforge voices or voices that are derived from voxforge voices to run, do I have to open source my application code as well as voices?

By Judy - 9/21/2009 - 1 Replies What is GPL and does the word "melarkia" have anything to do with it? I did a google search for "melarkia" and it sent me to VoxForge. Anyone know the answer? Thanks.

By kmaclean - 8/2/2009 - 1 Replies This article provides an support for VoxForge's use of the GPL: The different reasons for company code contributions. In it, the author states:

By vj61614 - 1/12/2009 pls help me in eliminating this error

By Anthony Martinez - 5/31/2008 - 1 Replies Hello, Congratulations for your project :) My question is simple, why not use also LGPL? If we could put this module in a commercial software, using LGPL would be mandatory to commit any modifications to the source code (that would bring more coders and investment into this module). This would be fair for the community, it would bring more people and more investment, and wouldn't be anti-enterprise. Would be the best of both worlds. Not all enterprises are evil. There are good enterprises that bring innovation, create jobs, pay taxes and make a good contribution to the world. The problem is when those companies become too greedy to share any innovation. Thats why LGPL is so nice. Think about it :) Cya

By Robert (Jamie) Munro - 1/27/2008 - 2 Replies Be very careful in the way you license this data. If you release a great collection of GPL data, and someone else releases a great collection of data under, for example, a CC-by license, it will probably be illegal to combine the two corpuses and make a working speech recognition product. I have been involved in 2 projects that went through immense pain because of this issue. The solution is as follows: Set up a proper legal entity to hold the data (an organisation). Make sure the organisation has good governance that end users will approve of. Make end users assign a non-exculsive irrevokeable perpetual license to the organisation for the organisation do to whatever it democratically decides in future. If you don't make this clear right from the start, any attempt to change the license, for whatever reason, will be impossible. For example, if a court rules that using a subset of the data derived from VoxForge on a hand-held device doesn't comply with the source code rule, and you have to bundle gigabytes of data in order to use VoxForge derived engines on mobile devices, you will be back to square one, and will have to start collecting data all over again. Who knows why a court may do that, but unless you can be sure, you can't risk not assigning data. The organisation could have in it's constitution, (terms of incorporation or whatever it's called in the relevant jurisdiction) that it will always make the data available under the GPL, but may additionally make it available under other licenses.

By Visitor - 10/10/2006 - 5 Replies I think there is a misunderstanding of the difference between audio data and source code. When I create an executable from source code, I might modify the source code. But when I create acoustic models, I don't modify the data that is used to train the acoustic models. So, if I have to distribute the data along with my models, I'd be distributing an identical copy of the data. (Not to mention the difficulties of distributing gigabytes of data...)