[Gzz-commits] gzz/Documentation/Manuscripts/Paper paper.tex

[Tuomas J. Lukka]

CVSROOT:        /cvsroot/gzz
Module name:    gzz
Changes by:     Tuomas J. Lukka <address@hidden>        02/12/01 07:07:17

Modified files:
        Documentation/Manuscripts/Paper: paper.tex 

Log message:
        twid

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/gzz/Documentation/Manuscripts/Paper/paper.tex.diff?tr1=1.170&tr2=1.171&r1=text&r2=text

Patches:
Index: gzz/Documentation/Manuscripts/Paper/paper.tex
diff -u gzz/Documentation/Manuscripts/Paper/paper.tex:1.170 
gzz/Documentation/Manuscripts/Paper/paper.tex:1.171
--- gzz/Documentation/Manuscripts/Paper/paper.tex:1.170 Sun Dec  1 07:04:33 2002
+++ gzz/Documentation/Manuscripts/Paper/paper.tex       Sun Dec  1 07:07:17 2002
@@ -376,6 +376,14 @@
 fashion\cite{rosenblatt62neurodynamics,widrow60adaptive}.
 This configuration is sometimes used in neural computation.
 
+The model explains readily why uniformly random texels (white noise)
+would not make easily distinguishable patterns: different instances
+of noise would all yield almost 
+exactly the same feature vector in brain:
+Noise has no global shape because there is no correlation between
+the random local features; it is simply perceived as the distribution
+of the local features, i.e., color and overall frequency 
+(the density of texels).
 
 From this 
 rough, qualitative model 
@@ -434,22 +442,13 @@
 
 In a sense, the model of perception should be {\em inverted}
 in order to produce a unique background from 
-a random vector seeded by the identity.
+a random vector.
 This type of approach has been used 
 by Ware and Knight\cite{ware95texture}, for inverting 
 the earliest stage of the visual system, 
 the spatial frequency detectors, in order to place 
 a particular vector or scalar field of data
 in the texture ``channel''.
-
-The model explains readily why uniformly random texels (white noise)
-would not make easily distinguishable patterns: different instances
-of noise would all yield almost 
-exactly the same feature vector in brain:
-Noise has no global shape because there is no correlation between
-the random local features; it is simply perceived as the distribution
-of the local features, i.e., color and overall frequency 
-(the density of texels).
 
 % To understand why it is possible to learn to discriminate particular
 % backgrounds easily, consider the task of learning {\em one} background 
texture.