[Gzz-commits] manuscripts/Paper paper.tex

[Janne V. Kujala]

CVSROOT:        /cvsroot/gzz
Module name:    manuscripts
Changes by:     Janne V. Kujala <address@hidden>        03/03/28 08:54:18

Modified files:
        Paper          : paper.tex 

Log message:
        reorg and some editing

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/manuscripts/Paper/paper.tex.diff?tr1=1.55&tr2=1.56&r1=text&r2=text

Patches:
Index: manuscripts/Paper/paper.tex
diff -u manuscripts/Paper/paper.tex:1.55 manuscripts/Paper/paper.tex:1.56
--- manuscripts/Paper/paper.tex:1.55    Fri Mar 28 07:53:07 2003
+++ manuscripts/Paper/paper.tex Fri Mar 28 08:54:17 2003
@@ -274,21 +274,6 @@
 %depends only on the values of its neighborhood (local characteristics).
 %XXX: resolution-dependency?
 
-%% XXX: this is not really texturing:
-There have been studies on 
-mapping texture appearance to an Euclidian texture space
-(see \cite{gurnsey01texturespace} and the references therein):
-in the reported experiments, three dimensions have been sufficient
-to explain most of the variation in the similarity judgements for
-artificial textures. 
-However, the texture stimuli have been somewhat simple 
-(no color, lack of frequency-band interaction, etc.).
-For some natural texture sets, 
-three dimensions have also been
-sufficient \cite{rao96texturenaming}, but often semantic connections cause the
-similarity to be context-dependant, making it hard to assess the 
-dimensionality.
-%% XXX: this is something we should experiment with our textures
 
 % In this article, we apply texture shading to synthesize a large number
 % of unique textures for distinguishing virtual objects.
@@ -561,7 +546,7 @@
 In these models, a bank of linear filters is applied to the texture followed
 by a nonlinearity and then another set of filters to extract features
 (see, e.g., Heeger\cite{heeger95pyramid}).
-There is also physiological evidence of filtering processes:
+There is also physiological evidence of the filtering processes:
 %The first stages 
 %of visual perception 
 %are fairly well known
@@ -612,7 +597,7 @@
 The simple model we use here assumes
 that at some point,
 the results from the different pre-attentive feature detectors,
-such as local and global shapes and colors, 
+such as different shapes and colors, 
 are combined to form an abstract \emph{feature vector}
 (see Fig.~\ref{fig-perceptual}).
 The feature vector is then used to compute 
@@ -631,10 +616,9 @@
 of noise would all yield 
 {\em almost 
 exactly the same feature vector} in the brain.
-Noise has no global shapes 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).
+Noise has no shape because there is no correlation between
+the local features; it is simply perceived as the distribution
+of texel colors and the overall frequency (the density of texels).
 
 From the 
 model 
@@ -653,7 +637,7 @@
 % - seeing different parts of the texture?
 % - ambiguous perception?
 \item The entropy of the feature vectors
-    over the distribution of textures, should be maximized.
+    over the distribution of textures should be maximized.
 \end{itemize}
 
 % which facilitates recognition and memorization of images.
@@ -673,7 +657,7 @@
 %We call this the principle of saving bits.
 
 The last part means essentially
-that if all square-like shapes were green, we would again be
+that if all square-like shapes were green, we would be
 wasting recognitive power.
 There should also be as many possible features in the distribution
     as possible. For example, if there were no yellow textures,
@@ -853,11 +837,12 @@
 them fairly close together around a uniformly chosen mean
 but also allows far-away hues
 in the same palette occasionally.
-The saturations are chosen from distribution emphasizing
+The saturations are chosen from a distribution emphasizing
 saturated colors: unsaturated colors can easily cause a too multicolored
 palette because the adaptive effects of the eye shift them
 towards the complementary colors of the more saturated colors
 in the palette.
+
 XXX: note: maximum saturation of light colors is limited
 
 % To produce a palette of
@@ -1410,6 +1395,21 @@
 and to make the ad hoc distributions more experimentally based.
 It could also be worthwhile to experiment with other ways
 of visualizing identity, such as unique edge shapes.
+
+There have been studies on 
+mapping texture appearance to an Euclidian texture space
+(see \cite{gurnsey01texturespace} and the references therein):
+in the reported experiments, three dimensions have been sufficient
+to explain most of the variation in the similarity judgements for
+artificial textures. 
+However, the texture stimuli have been somewhat simple 
+(no color, lack of frequency-band interaction, etc.).
+For some natural texture sets, 
+three dimensions have also been
+sufficient \cite{rao96texturenaming}, but often semantic connections cause the
+similarity to be context-dependant, making it hard to assess the 
+dimensionality.
+%% XXX: this is something we should experiment with our textures
 
 % The new graphics chips, ATI R300 and NVIDIA NV30 support
 % a great deal more of procedural texturing and it will