Newsgroups: comp.speech
Path: lyra.csx.cam.ac.uk!doc.ic.ac.uk!aixssc.uk.ibm.com!ibm.de!Munich.Germany.EU.net!Germany.EU.net!EU.net!uunet!almserv!dagger!g3urbp
From: g3urbp@fnma.COM (Bob Posey)
Subject: voice (not speech) recognition algo needed
Message-ID: <1994Aug2.154027.2208@almserv.uucp>
Sender: usenet@almserv.uucp
Nntp-Posting-Host: dagger
Reply-To: g3urbp@fnma.COM
Organization: Fannie Mae
Date: Tue, 2 Aug 1994 15:40:27 GMT
Lines: 18

Given: 1. Pre-recorded voice file of John Doe speaking his name ("John Doe").
       2. Newly recorded voice file of someone alleged to be the same John Doe, 
          also speaking his name ("John Doe").

Needed: To compare the two files to see if person 2 is the real John Doe.

I am an experienced Unix/C programmer, but new to voice processing.  Any help would
be greatly appreciated in these areas:

   1. Is this currently "doable"?

   2. Is what I want to do called "voice recognition" as opposed to "speech 
      recognition", since I don't need to have the code "understand" the words,
      only compare 2 voice files, of known content, for a match by speaker?

   3. Who, if anybody, sells libraries that do this (for SCO Unix).
   
Thanks very much to anyone who woul
