Newsgroups: comp.speech
Path: pavo.csi.cam.ac.uk!doc.ic.ac.uk!agate!howland.reston.ans.net!usc!news.service.uci.edu!ttinews!calvin.tti.com!cole
From: cole@soldev.tti.com (Randy Cole)
Subject: Re: a speaker's emotional state
Message-ID: <cole.313.748905316@soldev.tti.com>
Lines: 25
Sender: usenet@ttinews.tti.com (Usenet Admin)
Nntp-Posting-Host: calvin.tti.com
Organization: Citicorp-TTI
References: <27vikt$p02@uxa.ecn.bgu.edu>
Date: Fri, 24 Sep 1993 21:15:16 GMT

In article <27vikt$p02@uxa.ecn.bgu.edu> mfabw@uxa.ecn.bgu.edu (Adam B. Webber) writes:
>From: mfabw@uxa.ecn.bgu.edu (Adam B. Webber)
>Subject: a speaker's emotional state
>Date: 24 Sep 1993 14:43:25 -0500
>What can be deduced automatically about the emotional state
>of a speaker?  I'd appreciate any pointers to relevant work.
>
>Adam Webber
>AB-Webber@bgu.edu
>

Perhaps the leading example of quackery in the field of speech
was the Voice Stress Analyzer, which was sold as a lie detector.
The device supposedly determined whether or not a subject was
telling the truth by measuring the period-to-period variance of
voiced pitch.  Allegedly, stress results in tightening of the
muscles, including those involved with speech production, and
causes the pitch to vary less from pulse to pulse.  A relaxed
person supposedly allows things to flop around and causes
the pitch periods to vary more from pulse to pulse.

Good luck,

Randy Cole
cole@soldev.tti.com
