On 12/2/2014 3:08 PM, charles meyer wrote:
Hi Tom,
I spoke with someone who uses Linux and they shared that Sphinx -
voice translation is new so many may not have tried it yet.
Sorry for jumping the gun, so to speak.
It's speech recognition, not voice translation. If I throw you off the
top of the building, I'm going to hear your voice. If I push you near
the edge of the roof, I would hear your speech..
Sphinx has been around for at least 15 years in different forms. It has
been, and probably always will be a system designed for IVR (interactive
voice response, "speak or press one to get ignored by a customer service
representative"). It is not and never will be a system for
general-purpose speech recognition.
The only useful speech recognition packages are NaturallySpeaking with a
not very close runner-up of Windows speech recognition. Google speech
recognition would be in the running if it wasn't bound to a very limited
number of apps with no user accessible grammars. I'm currently
experimenting with running Windows in a KVM virtual machine, running
NaturallySpeaking there and find a way to see the output of
NaturallySpeaking back to the Linux host OS. If I can get the audio
stream clean enough, it looks like a promising technique for adding
speech recognition to Linux
Now all I need some help to figure out what I don't know about injecting
keystrokes into linux and may be help with fixing up KVM so it passes
audio cleanly under most conditions.
--
To unsubscribe e-mail to: users+unsubscr...@global.libreoffice.org
Problems? http://www.libreoffice.org/get-help/mailing-lists/how-to-unsubscribe/
Posting guidelines + more: http://wiki.documentfoundation.org/Netiquette
List archive: http://listarchives.libreoffice.org/global/users/
All messages sent to this list will be publicly archived and cannot be deleted