/[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 1.44 by jvk, Thu Mar 20 16:59:29 2003 UTC revision 1.45 by jvk, Thu Mar 20 17:59:01 2003 UTC
# Line 65  We apply a rough, qualitative model of v Line 65  We apply a rough, qualitative model of v
65  to motivate  to motivate
66  general principles for designing recognizably unique textures  general principles for designing recognizably unique textures
67  for use as backgrounds for data.  for use as backgrounds for data.
68  To be reconizable,  To be recongizable,
69  the texture should produce a random feature vector in the brain  the texture should produce a random feature vector in the brain
70  {\em after} visual feature extraction.  {\em after} visual feature extraction.
71    
# Line 301  on the $N$th-order statistics of texture Line 301  on the $N$th-order statistics of texture
301  placed (translated) $N$-gon for all different $N$-gons).  placed (translated) $N$-gon for all different $N$-gons).
302  However, the order of similarity in the statistics did not  However, the order of similarity in the statistics did not
303  consistently explain discrimination performance, and certain  consistently explain discrimination performance, and certain
304  distinctive local features were conjectured.  pre-attentive local features were conjectured.
305    
306  Julesz\cite{julesz81textons} proposed that texture discrimination could be  Julesz\cite{julesz81textons} proposed that texture discrimination could be
307  explained by the densities of textons, fundamental texture elements, such as  explained by the densities of textons, fundamental texture elements, such as
# Line 311  However, the textons are hard to define Line 311  However, the textons are hard to define
311  Much simpler filtering-based models can explain texture discrimination  Much simpler filtering-based models can explain texture discrimination
312  just as well \cite{bergen88earlyvision}.  just as well \cite{bergen88earlyvision}.
313  In these models, a bank of linear filters is applied to the texture followed  In these models, a bank of linear filters is applied to the texture followed
314  by a nonlinearity and then another set of filters to extract densities  by a nonlinearity and then another set of filters to extract features
315  of features (see, e.g., \cite{heeger95pyramid} for an application).  (see, e.g., \cite{heeger95pyramid} for an application).
316  %In \cite{heeger95pyramid}, new textures with appearance similar  %In \cite{heeger95pyramid}, new textures with appearance similar
317  %to a given texture are created by matching certain histograms  %to a given texture are created by matching certain histograms
318  %of filter responses.  %of filter responses.
319    
320    XXX: higher-level pre-attetive processes?
321    
322  %XXX: texture perception reviews  %XXX: texture perception reviews
323    
324  There have been studies on  There have been studies on
# Line 570  about recognizability. Line 572  about recognizability.
572    
573  The simple model we use here assumes  The simple model we use here assumes
574  that at some point,  that at some point,
575  the results from the  different feature detectors,  the results from the different pre-attentive feature detectors,
576  such as local and global shapes and colors,  such as local and global shapes and colors,
577  are combined to form an abstract \emph{feature vector}  are combined to form an abstract \emph{feature vector}
578  (see Fig.~\ref{fig-perceptual}).  (see Fig.~\ref{fig-perceptual}).

Legend:
Removed from v.1.44  
changed lines
  Added in v.1.45

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