Subject: | |
From: | |
Reply To: | |
Date: | Wed, 13 Aug 2008 08:58:14 -0500 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
On Wed, Aug 13, 2008 at 6:38 AM, Cleveland, Kyle E.
<[log in to unmask]> wrote:
> Lisa & Co.
>
> I work in the field of Emergency Management, and though my job is
> specifically to ensure that "government" continues to work in the event
> of a disaster (e.g., people stay in jail, taxes get colleted, waterworks
> stay running, etc.), my colleagues are looking at accessible web
> technologies to send email and other information interactively on the
> 'net. Do the current speech recognition web software packages currently
> being marketed really work well? What can you tell me about your
> experiences?
>
Kyle, I've worked with some of these technologies over the years.
You've probably run into the "state of the art" for current non
trained voice recognition if you've had to call any large companies
support line. If you're doing centralized servers these same
technologies can be used over the web. If you're doing client based
speech recognition the issues are rather different as at that point
the software can be "trained" for a particular user and have a much
wider vocabulary. Interesting side note: 1-800-goog-411 is Google's
voice recognition based system for free Business number look up. The
rumor is that they are using this as a test bed for training systems
for more general voice query (eg. search) services via mobile phones.
If you really want to look into this there are lots of places to look
but it might help to know more specifics...
--
Peter Hunsberger
-----------------------
To change your mail settings or leave the C-PALSY list, go here:
http://listserv.icors.org/SCRIPTS/WA-ICORS.EXE?SUBED1=c-palsy
|
|
|