Newsgroups: comp.speech
Path: pavo.csi.cam.ac.uk!pipex!uunet!europa.asd.contel.com!darwin.sura.net!zaphod.mps.ohio-state.edu!menudo.uh.edu!nuchat!texhrc!texhrc!ak45ldp
From: someone@Texaco.com (Larry D. Pyeatt)
Subject: Re: Phonemic analyzer construction
Message-ID: <1992Nov23.160224.16314@texhrc.uucp>
Sender: news@texhrc.uucp
Nntp-Posting-Host: microvax
Organization: Texaco
References:  <1992Nov23.133836.11680@hubcap.clemson.edu>
Date: Mon, 23 Nov 1992 16:02:24 GMT
Lines: 45

In article <1992Nov23.133836.11680@hubcap.clemson.edu>, bdbryan@eng.clemson.edu (Ben Bryant) writes:
|> G'day Sirs,
|> I am thinking about building a connectionist phoneme analyzer, and am interested
|> in finding out some ideas about how to go about designing the "higher-level"
|> classifier which will discriminate among the outputs from several previously
|> trained "subclass instant" neural nets.
|> 
|> Basically, the way this would work is that a suitable NN architecture would
|> be chosen for the "lower-level" signal analysis stage, and instances of this
|> architecture would be trained using TIMIT or some other large database.
|> 
|> The way the training would take place is as follows:
|> 1) first the training tokens for each phonemic subclass would be extracted
|>    from the database.
|> 2) the phoneme tokens for each phonemic subclass extracted in step one would
|>    then be preprocessed with an appropriate feature representation technique.
|> 3) network instances would be trained using the chosen neural network architecture.
|>    A network instance will be trained for each phonemic subclass (i.e., voiced-stops,
|>    unvoiced-stops, diphthongs, vowels, etc.).
|> 4) after training all network instances, the outputs from the trained subnetworks
|>    would "somehow" be arbitrated to provide a decision of which phoneme was uttered
|>    within a given region of signal.
|> 
|> -The "somehow" in step 4) is what I really could use some help with. Any other
|> ideas for this system would be welcome as well. Thank you very much.
|> 
|> Sincerely,
|> -Benjamin Bryant 
|> 
|> 
|> 

I would suggest using a BlackBoard system to do the arbitration.
You may want to look at HearSay II for inspiration.  Alternatively,
you could try a connectionist approach to the arbitration problem.
I do not know of anyone who has done previous work with the 
connectionist approach, but would like to hear any ideas.  I have
been thinking about doing something similar to what you have
suggested for my PhD research, which I hope to start next fall.

-- 
Larry D. Pyeatt                 The views expressed here are not
Internet : pyeatt@texaco.com    those of my employer or of anyone
Voice    : (713) 975-4056       that I know of with the possible
                                exception of myself.
