Newsgroups: comp.ai.neural-nets
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!howland.reston.ans.net!pipex!uunet!spstimes.sps.mot.com!mogate!newsgate!ninja!bobse
From: bobse@ninja (Robert W. Seaton Jr.)
Subject: Selective Attention
Message-ID: <1994Nov11.215832.23142@newsgate.sps.mot.com>
Sender: news@newsgate.sps.mot.com
Nntp-Posting-Host: ninja.sps.mot.com
Organization: sps
X-Newsreader: TIN [version 1.2 PL0]
Date: Fri, 11 Nov 1994 21:58:32 GMT
Lines: 20

John,

Gail Carpenter (one of the principle developers of ART[123]) talks ALOT
about selective attention. In fact, that is what the feedback units are
doing in ART (as per my limited understanding) providing a category back
to the input layer so that it can focus selective attention on a particular
set of inputs. So it might be helpful to check out any references on using
ART in speech recognition (sorry, I don't have any refs).

:) bobse


--
+=========================================================================+
| Bob Seaton                   External: Motorola ECO                     |
| bobse@swarm.sps.mot.com                5918 W. Courtyard Drive, Ste 200 |
| Parallel Stuff                         Austin, TX 78730                 |
| Phone: (512) 794-4069        Internal: TX73-F30                         |
| FAX:   (512) 794-4001                                                   |
+=========================================================================+
