Newsgroups: comp.speech
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!ix.netcom.com!netcom.com!alvin
From: alvin@netcom.com (Alvin H. White)
Subject: Speech Recognition Demonstration
Message-ID: <alvinCynIxJ.EJ4@netcom.com>
Organization: NETCOM On-line Communication Services (408 261-4700 guest)
Date: Wed, 2 Nov 1994 17:51:18 GMT
Lines: 59


ajr@eng.cam.ac.uk (Tony Robinson) writes:

>Cambridge University connectionist speech recognition group will be
>demonstrating their current work aimed at the automatic recognition of
>continuously spoken, very large vocabulary English speech.  We will be
>available to talk about the latest speech recognition systems, to
>explain how they work and how well they perform.  This will be backed up
>by an open demonstration whereby all who are interested can try out the
>system to explore the strengths and weaknesses of this state of the art
>research system.

>The demonstration will be held from 2pm to 4pm on Friday, 4 November in
>the ECR laboratory on the second floor of the Inglis building in the
>Engineering department.  All interested are welcome to attend.


>Tony Robinson

It sounds wonderful. Unfortunately, or fortunately, I am almost exactly
1/3 of the way around the planet and with the price of airplanes and
gasoline (petrol?) being what it is ...

A couple of weeks ago I sort of took part in some online teleconference
regarding Friends of the National Library of Medicine out of Washington,
D.C. Tony, I don't know if you all have considered focusing world 
attention of your subject by means of such like but it seems one of the
latest rages and means to escape playing solitaire on one's Internet
computer.

What would one call it, "A Tony Cocktail Party" or "The Cambridge A-Go-Go."
Well, I just wanted you to know that I was thinking about you.

It seems that if we all had Ora's gizmo and some of that Joe Campbell
CELP stuff we could all sing and then have the Cambridge Machine make
us all sound like qualified applicants of the "Pavarotti and the Angelic
Choir World Singing Olympiad." That is a thought for future years after
watching Andrew Hunt's comp.speech FAQ go by every few weeks. I read it
a bit from time to time. Thank you Andrew. 

I think I am beginning to think like this is "Internet Talk Radio."
Each morning I get up, fire up the machine, download the day's MIDI music
files broadcast, get the music going and then...

Speaking of that, somebody could do a PR gimick, like Cambridge, were
we at home could sing-a-long and then we could submit our homemade 
vocals to sync with the broadcast MIDI.

Then Tony could try to recognize what was said and who ever got the
highest recognition score could have their submission broadcast
on the net as the best example to imitate if you wanted your speech
to be understood by a computer.

Good luck with your open house. If this becomes an annual event and I
am ever in your neck of the woods at the right time I hope to drop by.

-- 

alvin@netcom.com
