Newsgroups: comp.speech
Path: pavo.csi.cam.ac.uk!warwick!zaphod.crihan.fr!vishnu.jussieu.fr!univ-lyon1.fr!scsing.switch.ch!xlink.net!howland.reston.ans.net!ux1.cso.uiuc.edu!sdd.hp.com!nigel.msen.com!heifetz!agog!greg
From: greg@agog.com (Greg Cockroft)
Subject: Re: Help: a simple speech pattern matching?
Message-ID: <1993Jul14.131740.1093@agog.com>
Sender: greg@agog.com (Greg Cockroft)
Organization: Agog, Inc.
References: <21vgi4$v39@msuinfo.cl.msu.edu>
Date: Wed, 14 Jul 1993 13:17:40 GMT
Lines: 30

In article <21vgi4$v39@msuinfo.cl.msu.edu> leeda@cps.msu.edu (Dae Hee Lee)  
writes:
> I have implemented a simple speech pattern matching, but failed. I don't
> know what's wrong.  The process was follows:
> I recorded hello.wav and goodbye.wav with 11k and 16-bit.  As I
> extracted25-channel lpc for each frame.  (I used the frame size 256 and
> window size 512.)  I saved in a tablet the parameters frame by frame.
> Later I recored another sound file newgoodbye.wav, and processed the
> sound in the same way as before and extracted the parameters, and tried
> to match it with the hello and goodbye by means of dynamic time warping
> (nonlinear method).  But it was misclassified.  So I changed the frame
> size and window size, or extracted other parameters such as lpc ceps,
> fft ceps, etc. instead, but the result was the same.
> Could you point out what is wrong?
> Thanks!
> 
> DH
> 
> 

A good debugging method is to print out all the distances of the entire  
array normalized by the number of distance tests that have been made.
Run the same file against itself. You should get a diagonal line of zeros.

	-greg.
-- 
Greg Cockroft                       greg@agog.com    
Agog, Inc.                          (517) 627-2186   fax 2673
13780 North River Highway
Grand Ledge, MI  48837
