Re: [g-a-devel] speech recognition and accessibility

[Willie Walker <William Walker Sun COM>]

Hey All:

Just a couple comments...

I believe Speech2Text is using the Julius
(http://julius.sourceforge.jp/en_index.php?q=en/index.html) engine,
which seems to be a decent system geared towards dictation tasks.  

An interesting and related project is VoxForge
(http://www.voxforge.org/), which is a system for gathering training
data and making it available for people to train acoustic models.  IMO,
this kind of work is part of what is dearly needed to help open source
speech recognition work.  Cleaning up the data, transcribing it, and
then building acoustic models from it is another thing... :-)

Nowadays, I happen to lead the Orca project.  In a recent past life,
however, I led the development of the Sphinx-4 speech recognition
system, which is still under development by some of the original members
of the team.  I also was the Principal Investigator for Speech in Sun
Labs for several years, doing work both on understanding the user aspect
of speech systems as well as the technical side of speech engines
themselves.  I'd be happy to engage you in a discussion about various
approaches to this problem.

Will

On Sat, 2007-06-09 at 11:18 +0200, Henrik Nilsen Omma wrote:
> Nickolay V. Shmyrev wrote:
> > В Птн, 08/06/2007 в 23:26 -0700, Peter Korn пишет:
> >   
> >> Hi Nickolay,
> >>
> >> I agree with Malte - this technology has much wider application than 
> >> just for those interested in GOK; it probably makes more sense to keep 
> >> it separate.
> >>     
> >
> > Heh, I was too fast writing that. I completely forgot that the code is
> > very old and rather outdated nowadays. And there is not so much code to
> > share - only gok-spy actually. Now the only concern is the speed of
> > inclusion but its a minor issue.
> >
> >   
> >> What are your long term plans for it?  Do you envision this remaining a 
> >> command-and-control application, or do you want to also include 
> >> dictation?  Do you have a long term plan to keep working on this?
> >>     
> >
> > Since so many people are interested in dictation, it will be supported.
> > I'm afraid current state of art in SR doesn't allow perfect dictation
> > but lets hope situation will improve one day. We can also allow
> > by-letter dictation of first stages.
> >
> >   
> >> Also, do you have any users with physical impairments involved in the 
> >> project?
> >>     
> >
> > We'd certainly like to see someone involved.
> 
> You should contact this group 
> http://sourceforge.net/projects/speech2text/ (Peter on CC)
> 
> They are working on a similar project with Qt.
> 
> Henrik
> _______________________________________________
> Gnome-accessibility-devel mailing list
> Gnome-accessibility-devel gnome org
> http://mail.gnome.org/mailman/listinfo/gnome-accessibility-devel