doc: improves style, activates syntax highlighting

docsrc/aib.html

@@ -24,21 +24,21 @@
 <p>We quantize this data on a fixed lattice (a 20x20 grid shown in the
 figures below), and construct histograms for each class.</p>
 
-<pre>
+<precode type='matlab'>
 f1 = quantize(X1,D,K) ;
 f2 = quantize(X2,D,K) ;
 f3 = quantize(X3,D,K) ;
 
 Pcx(1,:) = vl_binsum(Pcx(1,:), ones(size(f1)), f1) ;
 Pcx(2,:) = vl_binsum(Pcx(2,:), ones(size(f2)), f2) ;
 Pcx(3,:) = vl_binsum(Pcx(3,:), ones(size(f3)), f3) ;
-</pre>
+</precode>
 
 <p>Next we apply AIB:</p>
 
-<pre>
+<precode type='matlab'>
 [parents, cost] = vl_aib(Pcx) ;
-</pre>
+</precode>
 
 <p>This provides us with a list of parents of each column
 in <code>Pcx</code>, forming a tree of merges. We can now "cut" this

docsrc/compiling.html

@@ -18,16 +18,16 @@ <h1 id="compiling.unix">Compiling on Unix</h1>
 standard GNU toolchain. It is usually sufficient to change to VLFeat
 root directory, denoted <code>VLFEATROOT</code> in the following, and
 type <code>make</code>:</p>
-<pre>
+<precode type='sh'>
 > cd VLFEATROOT
 > make
-</pre>
+</precode>
 <p>The make script attempts to automatically detect the host
 architecture and configure itself accordingly. If the architecture is
 not detected correctly, it can be specified manually. For instance</p>
-<pre>
+<precode type='sh'>
 > make ARCH=glx
-</pre>
+</precode>
 <p>compiles for GNU/Linux 32-bit. <code>make help</code> can be used
 to obtain a list of other useful options.</p>
 
@@ -40,9 +40,9 @@ <h1 id="compiling.unix">Compiling on Unix</h1>
 (e.g. <code>/Applications/MATLAB_R2009b.app/</code>). The <code>mex</code>
 script is usually located in <code>MALTABROOT/bin/mex</code>. Then run
 the compilation with the command</p>
-<pre>
+<precode type='sh'>
 > make MEX=MATLABROOT/bin/mex
-</pre>
+</precode>
 <p>VLFeat must be compiled for the architecture used by MATLAB (run
  MATLAB's <code>computer</code> command to obtain this
  information). On Mac OS X it is often necessary to turn on 64-bit
@@ -56,29 +56,29 @@ <h1 id="compiling.unix">Compiling on Unix</h1>
 compiling for this platform, but the architecture must be specified
 manually when compiling, either like:</p>
 
-<pre>
+<precode type='sh'>
 make ARCH=maci64
-</pre>
+</precode>
 
 <p>or</p>
 
-<pre>
+<precode type='sh'>
 make ARCH=maci
-</pre>
+</precode>
 
 <p>Unfortunately, MATLAB 2009b's <code>mex</code> script has a bug that
 must be manually fixed in order for this procedure to work
 correctly. It is recommend to make backup of the <code>mex</code>
 script before attempting this. Th fix the bug, edit the
 <code>MATLABROOT/bin/mex</code> script by cutting the line:</p>
 
-<pre>
+<precode type='sh'>
 get_arch    # Call $MATLAB/bin/util/arch.sh
-</pre>
+</precode>
 
 <p>and pasting it after the processing of options:</p>
 
-<pre>
+<precode type='sh'>
 ...
 shift
 done
@@ -89,7 +89,7 @@ <h1 id="compiling.unix">Compiling on Unix</h1>
 if [ $v7_compat -eq 1 ] ; then
 fc_flags="$fc_flags -DMX_COMPAT_32"
 ...
-</pre>
+</precode>
 </div>
 
 <h1 id="compiling.windows">Compiling on Windows</h1>

docsrc/doc.html

@@ -2,29 +2,37 @@
 
 <group>
 
-<p>The VLFeat documentation has three parts:</p>
-
-<ul>
-
-<li><p><a href="%pathto:mdoc;">Matlab help.</a>  VLFeat Matlab
-commands embed documentation that can be viewed with the
-Matlab <code>help</code> command (for instance, type <code>help
-vl_sift</code> to read the <code>vl_sift</code> command
-documentation). For convenience, the same documentation is also
-available in <a href="%pathto:mdoc;">HTML format.</a></p></li>
-
-<li><p><a href="%pathto:api;">C API.</a>  VLFeat source
-code embeds extensive documentation. This includes <b>descriptions of
-all algorithms</b> which can be useful even if you do not plan to use
-the C library directly. </p>
-</li>
-
-<li><p><a href="%pathto:man;">Man pages.</a> The command line utilities
-bundled with VLFeat are described by Unix man pages. These
-are found in the <code>src/</code> subdirectory, or can be viewed in
- <a href="%pathto:man;">HTML format.</a></p></li>
-
-</ul>
+<p>The VLFeat <em>reference documentation</em> has three parts:</p>
+
+<table class="boxes">
+  <tr>
+    <td style="height:6em;padding-right:0.5em;width:33%;">
+      <div class="box">
+        <h1><a class="plain" href="%pathto:mdoc;">MATLAB functions</a></h1>
+        <p>The reference documentation of VLFeat MATLAB commands (this
+        is an on-line version of the documentation built-in the
+        command themsevles).</p>
+      </div>
+    </td>
+    <td style="height:6em;padding:0 0.5em;width:33%;">
+      <div class="box">
+        <h1><a class="plain" href="%pathto:api;">C API
+        reference</a></h1><p>VLFeat source code embeds extensive
+        documentation. This includes <b>descriptions of all
+        algorithms</b> which can be useful even if you do not plan to
+        use the C library directly.</p>
+      </div>
+    </td>
+    <td style="height:6em;padding-left:0.5em;width:33%;">
+      <div class="box">
+      <h1><a class="plain" href="%pathto:api;">Man pages</a></h1>
+      <p>The documentation of the command line
+      utilities bundled with VLFeat (this is an on-line version
+      of the Unix man found in the <code>src/</code>
+      subdirectory.</p></div>
+    </td>
+  </tr>
+</table>
 
 <p>In addition to the documentation, there are also 
 <a href="%pathto:tut;">tutorials</a> which introduce many of the

docsrc/download.html

@@ -6,6 +6,9 @@
 VLFeat, simply download and unpack the latest binary package and add
 the appropriate paths to your environment (see below for details).</p>
 
+<table class="boxes">
+<tr>
+<td style="height:6em;padding-right:0.5em;">
 <div class="box">
 <h1>Downloads</h1>
 <ul>
@@ -20,7 +23,8 @@ <h1>Downloads</h1>
  <li><a href="%pathto:root;download/">Previous versions ...</a></li>
 </ul>
 </div>
-
+</td>
+<td style="height:6em;padding-left:0.5em;">
 <div class="box">
 <h1>Install</h1>
 <ul>
@@ -29,6 +33,9 @@ <h1>Install</h1>
 <li><a href="%pathto:install.c;">C library</a></li>
 </ul>
 </div>
+</td>
+</tr>
+</table>
 
 <div class="clear">&nsbp;</div>
 
@@ -38,7 +45,7 @@ <h1>Repository access</h1>
 can <a href="http://github.com/vlfeat/vlfeat/tree/master">browse our
 Git repository</a> or download it by</p>
 
-<pre>git clone git://github.com/vlfeat/vlfeat.git</pre>
+<precode type='matlab'>git clone git://github.com/vlfeat/vlfeat.git</precode>
 
 <p>(This will require <a href="http://www.git-scm.org/">Git</a> to be
 installed). The top of the master branch corresponds to the most
@@ -53,7 +60,7 @@ <h1>Repository access</h1>
 like <code>git format-patch</code>:
 </p>
 
-<pre>git format-patch origin</pre>
+<precode type='matlab'>git format-patch origin</precode>
 
 <p>For major additions, we prefer to handle collaboration through
 <a href="http://github.com/">github</a>. Follow 

docsrc/doxygen.conf

@@ -700,7 +700,7 @@ FILTER_SOURCE_FILES    = NO
 # Note: To get rid of all source code in the generated output, make sure also
 # VERBATIM_HEADERS is set to NO.
 
-SOURCE_BROWSER         = YES
+SOURCE_BROWSER         = NO
 
 # Setting the INLINE_SOURCES tag to YES will include the body
 # of functions and classes directly in the documentation.
@@ -717,21 +717,21 @@ STRIP_CODE_COMMENTS    = YES
 # then for each documented function all documented
 # functions referencing it will be listed.
 
-REFERENCED_BY_RELATION = YES
+REFERENCED_BY_RELATION = NO
 
 # If the REFERENCES_RELATION tag is set to YES
 # then for each documented function all documented entities
 # called/used by that function will be listed.
 
-REFERENCES_RELATION    = YES
+REFERENCES_RELATION    = NO
 
 # If the REFERENCES_LINK_SOURCE tag is set to YES (the default)
 # and SOURCE_BROWSER tag is set to YES, then the hyperlinks from
 # functions in REFERENCES_RELATION and REFERENCED_BY_RELATION lists will
 # link to the source code.
 # Otherwise they will link to the documentation.
 
-REFERENCES_LINK_SOURCE = YES
+REFERENCES_LINK_SOURCE = NO
 
 # If the USE_HTAGS tag is set to YES then the references to source code
 # will point to the HTML generated by the htags(1) tool instead of doxygen
@@ -745,7 +745,7 @@ USE_HTAGS              = NO
 # will generate a verbatim copy of the header file for each class for
 # which an include is specified. Set to NO to disable this.
 
-VERBATIM_HEADERS       = YES
+VERBATIM_HEADERS       = NO
 
 #---------------------------------------------------------------------------
 # configuration options related to the alphabetical class index

docsrc/dsift.html

@@ -15,10 +15,10 @@ <h1 id="tut.dsift.fast">Dense SIFT as a faster SIFT</h1>
 
 <p>The main advantage of using <code>vl_dsift</code>
 over <code>vl_sift</code> is speed. To see this, load a test image</p>
-<pre>
+<precode type='matlab'>
 I = imread(fullfile(vl_root,'data','a.jpg')) ;
 I = single(vl_imdown(rgb2gray(I))) ;
-</pre>
+</precode>
 
 <p>To check the equivalence of <code>vl_disft</code>
 and <code>vl_sift</code> it is necessary to understand in detail how
@@ -46,7 +46,7 @@ <h1 id="tut.dsift.fast">Dense SIFT as a faster SIFT</h1>
 <p>Thus the following code produces equivalent descriptors using
   either DSIFT or SIFT:</p>
 
-<pre>
+<precode type='matlab'>
 binSize = 8 ;
 magnif = 3 ;
 Is = vl_imsmooth(I, sqrt((binSize/magnif)^2 - .25)) ;
@@ -55,7 +55,7 @@ <h1 id="tut.dsift.fast">Dense SIFT as a faster SIFT</h1>
 f(3,:) = binSize/magnif ;
 f(4,:) = 0 ;
 [f_, d_] = vl_sift(I, 'frames', f) ;
-</pre>
+</precode>
 
 <p>The difference, of course, is that DSIFT is much faster.</p>
 
@@ -89,10 +89,10 @@ <h1 id="tut.dsift.phow">PHOW descriptors</h1>
 <p>VLFeat includes a simple wrapper, <code>vl_phow</code>, that does
 exactly this:</p>
 
-<pre>
+<precode type='matlab'>
 im = imread(fullfile(vl_root, 'data', 'a.jpg')) ;
 [frames, descrs]=vl_phow(im2single(im)) ;
-</pre>
+</precode>
 
 <p>Note that this typically generate a very large number of features.
 In this example, there are 162,574 features.</p>