On Sun, Jun 11, 2000 at 04:47:43PM -0500,
David Champion <[EMAIL PROTECTED]> wrote:
> On 2000.06.11, in <005801bfd3a3$5bbb1320$f8a0a6d4@alderaan>,
> "Daniel González Gasull" <[EMAIL PROTECTED]> wrote:
> > Hello all.
> >
> > Perhaps this is a silly question, but, is there any
> > program (say catwav) which let me do something like
> >
> > catwav message.wav
> >
> > and display the message in my console with speech
> > recognition?
>
> I'd say "almost certainly not". As far as I know, speech recognition
> is not yet to the point that software can reliably do it without being
> "trained" to the voice that's producing the speech.
OK but, how it works without training? Perhaps
it works at 75% of the times. It can be enough for
me.
BTW, if I receive many voice mail from the same
people, I can train the speech recognition software
with those messages.
An idea: With mailcap we can have 2 entries:
audio/wav; catwav %s
audio/wav; catwav -dump %s; copiousoutput
The first is for training the software. The more I
train the software, the better it will work. The
second entry is for auto_view.
Notes: "catwav" is a ficticious speech recognition program.
"catwav -dump" works like "lynx -dump", you know ;-)
Is it still sci-fi?
> The best way to handle this would be to put the voice recognition
> software on the mobile phone, and have it submit the email as
> multipart/alternative. That's a suggestion for anyone out there in the
> mobile phone business.
Yes. It's a good suggestion.
--
Daniel González Gasull "Los hombres suelen
[EMAIL PROTECTED] _ _ confundir los límites de
PGP RSA key 1024/EEA93A69 .oooO / ) ( \ Oooo. su propia visión con los
A3 8C 93 C0 B8 47 78 0C ( ) / ( ) \ ( ) límites del mundo."
37 85 44 59 63 32 8A 68 \ ( ( ) ( ) ) / -- Schopenhauer
---\_).oooO-Oooo.(_/--
PGP signature
