Xref: utzoo comp.ai:5034 comp.misc:7322
Path: utzoo!utgpu!utstat!jarvis.csri.toronto.edu!rutgers!uwm.edu!cs.utexas.edu!rice!uw-beaver!fluke!ssc-vax!bcsaic!rwojcik
From: rwojcik@bcsaic.UUCP (Rick Wojcik)
Newsgroups: comp.ai,comp.misc
Subject: Re: Speech Recognition Breakthru
Message-ID: <16867@bcsaic.UUCP>
Date: 9 Nov 89 19:30:52 GMT
References: <2932@hub.UUCP>
Reply-To: rwojcik@bcsaic.UUCP (Rick Wojcik)
Organization: Boeing Computer Services AI Center, Seattle
Lines: 17

In article <2932@hub.UUCP> probert@aviary.gm.hac.com writes:
>On CNN this weekend was a short piece on a development in speech recognition
>by Roger Moore of the Royal Signals & Radar Establishment in Malvern England.
>It seems he has come up with a simple algorithm of transcribing speech.
>Can anyone direct me to recent papers by Dr. Moore or the work he is doing?

I am also interested in this story, although it should be treated with the
same credence as those similar stories that have appeared over the past three
decades.  From a linguistic point of view, it is impossible to come up with a
simple algorithm for transcribing speech.  The issues involved in that task
are horrendously complex.  What Moore has come up with is probably a new way
of segmenting the acoustic stream.  I'd like to know more about it.


-- 
Rick Wojcik   csnet:  rwojcik@atc.boeing.com	   
              uucp:   uw-beaver!bcsaic!rwojcik