/[gzz]/manuscripts/Paper/paper.tex
ViewVC logotype

Diff of /manuscripts/Paper/paper.tex

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.80 by tjl, Mon Mar 31 10:32:22 2003 UTC revision 1.81 by tjl, Mon Mar 31 10:33:15 2003 UTC
# Line 299  some theories Line 299  some theories
299  assume certain primitive shapes whose  assume certain primitive shapes whose
300  structure facilitates recognition.  structure facilitates recognition.
301    
 The simple model we use here assumes  
 that at some point,  
 the results from the different pre-attentive feature detectors,  
 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  
 which concept the particular  
 input corresponds to by comparing it to memorized models  
 in a simple perceptron-like  
 fashion\cite{rosenblatt62neurodynamics,widrow60adaptive}.  
 This configuration is commonly used in neural computation.  
   
302    
303  % In this article, we apply texture shading to synthesize a large number  % In this article, we apply texture shading to synthesize a large number
304  % of unique textures for distinguishing virtual objects.  % of unique textures for distinguishing virtual objects.
# Line 575  visual system. Line 562  visual system.
562  %providing an infinite source of unique backgrounds.  %providing an infinite source of unique backgrounds.
563  %generating textures based on seed numbers [identity]  %generating textures based on seed numbers [identity]
564    
565    The simple model we use here assumes
566    that at some point,
567    the results from the different pre-attentive feature detectors,
568    such as different shapes and colors,
569    are combined to form an abstract \emph{feature vector}
570    (see Fig.~\ref{fig-perceptual}).
571    The feature vector is then used to compute
572    which concept the particular
573    input corresponds to by comparing it to memorized models
574    in a simple perceptron-like
575    fashion\cite{rosenblatt62neurodynamics,widrow60adaptive}.
576    This configuration is commonly used in neural computation.
577    
578  \begin{figure}  \begin{figure}
579  %\fbox{\vbox{\vskip 3in}}  %\fbox{\vbox{\vskip 3in}}
580  \ifpics  \ifpics
# Line 604  with maximum entropy is... Line 604  with maximum entropy is...
604  % The basic assumption of the model is that an image  % The basic assumption of the model is that an image
605  % is perceived as a set of features  % is perceived as a set of features
606    
607  This  This rough, qualitative
 rough, qualitative  
608  model is able to explain why uniformly random texels  model is able to explain why uniformly random texels
609  do not make easily distinguishable background textures:  do not make easily distinguishable background textures:
610  after the ``pre-processing'',  after the ``pre-processing'',

Legend:
Removed from v.1.80  
changed lines
  Added in v.1.81

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26