Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!uunet!mcsun!hp4nl!media01!pkr
From: pkr@media01.UUCP (Peter Kriens)
Newsgroups: comp.ai.neural-nets
Subject: Fonts, kerning and neural nets
Message-ID: <2163@media01.UUCP>
Date: 12 Mar 91 21:21:49 GMT
Reply-To: pkr@media01.UUCP (Peter Kriens)
Organization: aQute, Netherlands
Lines: 18

I am very new to neural nets, but what I have read about it, it could be
very helpfull with a problem I have got. I am working on a project
where we have to find the optimal optical spacing between character pairs.
For example, the adjacent V/A pair should be a little bit compacted,
called kerning. Per font there are about a 1000 pairs which need kerning
and there are a few thousand fonts.

For many fonts, there are kerning tables that could be used as learning
input's. My problems:

	a:Is this indeed a problem that could be handled by neural nets
	b: I really don't know where to start, any pointers?
	c: Is there some public domain program which I can use (C?)

Thanks for the time.

Peter Kriens
pkr@media01