[sf-lug] linux voice recognition?

[Rigel Christian]

Does anyone have experience with rolling their own voice recognition using
CMU's Sphinx-3 or Sphinx-4? or others, too, uts just sphinx looks like the
most mature project.

I wanted to learn more about it for use in a medical transcription context,
but the introductory literature seems really kind of impenetrable to someone
who knows squat about the field.

since the production of acoustic models seems to depend on a significant
number of people reading the same thing into a microphone, i would also
probably want to try and shanghai other members of the group into
participating, once i figured out more details.

cheers.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://linuxmafia.com/pipermail/sf-lug/attachments/20080919/f2299ab2/attachment.html>