Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!usc!apple!motcsd!hpda!hpcuhb!hpcuha!hp-ptp!garye
From: garye@hp-ptp.HP.COM (Gary_Ericson)
Newsgroups: comp.cog-eng
Subject: State of handwriting recognition
Message-ID: <1440006@hp-ptp.HP.COM>
Date: 2 Oct 89 21:52:36 GMT
Organization: HP Pacific Technology Park - Sunnyvale, Ca.
Lines: 20

In the 1987 CHI Conference, there was a panel discussion entitled, "Issues
Limiting the Acceptance of User Interfaces Using Gesture Input and Handwriting
Character Recognition".  The conference proceedings gave a synopsis of the
stance of each panel member (I didn't attend so I didn't hear the actual
discussion).  They each felt that handwriting recognition had been given a bad
name, mostly due to shortfalls that technology has had in the past.  But they
believed that handwriting recognition was much better now, enough so that
commercial products were definitely feasible, both in technology and cost.  The
only problem is that users and developers don't *believe* that it's feasible.

Did anyone attend this panel discussion?  Where these the opinions of people
with a vested interest (some were from companies making h.r. software), or were
they on-the-mark with their comments?  Is h.r. really a viable input mechanism
(in applications where it makes sense), i.e., can I find a piece of software,
or software/hardware mix, that will recognize 2-3 characters per second with 
99% accuracy?  Would that be upper-case print only, upper/lower-case print, or
cursive script?

Gary Ericson - System Engineering, Workstation Systems Division
               phone: (408)746-5098  mailstop: 101N  email: gary@hpdsla9.hp.com