[OFF-TOPIC] Re: ideological speed bumps

[Bill Cox]

I'm also following this thread.  I had to program by voice for three
years in the '90s, first with Dragon Dictate, and then with Naturally
Speaking.  I eventually wrote 1,600 voice macros mostly to control
emacs to help me do my job.

When I started with Dragon Dictate, I was excited about the rapid
progress for the disabled.  Dragon Systems was doing wonderful things
for us.  Then, Dragon Systems shipped a tool for voice-dictation aimed
at regular users.  Progress stopped, almost dead right then, and never
picked up again.

I want to add voice recognition solutions to Vinux, which is built on
Ubuntu Lucid.  However, Naturally Speaking remains the best voice
recognition engine, and there's little reason to believe the recent
owners, Nuance, will port it.  Nuance also bought Eloquence, the best
TTS engine for the blind, IMO, since it can be well understood at very
high speeds.  Eloquence use to run on Linux, but there is no evidence
that Nuance will release any new version for our platform.

Modern open-source research and advancement is somewhat promising.
Espeak seems to get better each year, though it's far behind Eloquence
for high speed.  Then there's svox pico around the corner from Google,
which may help bring open-source natural voices along.  On the
recognition side, there's some advancement, but I have yet to see any
good FOSS demo on Linux.

One dumb thought I had this morning: Could we just call the original
developers and ask for their help as consultants on FOSS ASR and TTS?
They must be long gone from their companies, and I imagine that their
non-competes have expired.  What really counts is the know-how.  If
they could consult on algorithm specification and development, without
giving up any trade-secrets, they wouldn't have to write one line of
code.  I'd be we'd find FOSS devs willing to code it up.

Bill

On Sun, May 16, 2010 at 9:51 AM, Pia <pmikeal at comcast.net> wrote:
> I just wanted to ask that you guys not take this topic off list.  It was
> one of the most seriously useful conversations that has been on here for a
> long time, because it looks at the future of a barely functional state of
> things which is really what we all should be concerned about.  So, I have
> been reading the thread closely.  I just have not added much yet, because
> I would just be repeating much of what has already be said at this point.
>
> Kind Regards,
>
> Pia
>
>
> --
> Ubuntu-accessibility mailing list
> Ubuntu-accessibility at lists.ubuntu.com
> https://lists.ubuntu.com/mailman/listinfo/ubuntu-accessibility
>