Newsgroups: comp.speech
Path: pavo.csi.cam.ac.uk!pipex!uunet!munnari.oz.au!uniwa!swanee.ee.uwa.oz.au!hera!mylene
From: mylene@swanee.ee.uwa.edu.au (Mylene Pijpers)
Subject: isolated digits databases
Message-ID: <mylene.718354488@swanee>
Sender: news@swanee.ee.uwa.oz.au
Organization: Elec Eng, Univ of Western Australia
Date:  6 Oct 92 06:54:48 GMT
Lines: 36



Hi,

We are working on speaker-independent recognition of isolated digits.

Because we reached the point where improvements in recognition results 
might either be due to an improvent of the system, or due to an
insufficiently large data~set we are currently looking for large 
speech~databases of isolated digits.
Ideally we need some 1000 speakers to avoid speaker related overtraining.
Somehow it seems there are no such databases around, at least we have not
been able to locate them.

So here is the urgent question:

Any, I repeat, ANY leads to the whereabouts of large databases of
isolated digits are welcome at this e-mail address.


I promise to post a summary. 
And it might be a subject for a Frequently Asked Questions file.


Mylene





Myl\`ene Pijpers    --    mylene@swanee.ee.uwa.edu.au    |   _--_|\
Centre for Intelligent Information Processing Systems    |  /ciips \ 
The University of Western Australia                      |  *_.--._/    
NEDLANDS WA 6009 Australia      Phone: +61-9-380-3897    |        v 
                                                    

