*VERSION 3.25 OF TOOLBOX*

qinhongwei · Dec 16, 2013 · 1aa5c4f · 1aa5c4f
1 parent 066a755
commit 1aa5c4f
Show file tree

Hide file tree

Showing 9 changed files with 32 additions and 23 deletions.
diff --git a/doc/channels/Contents.html b/doc/channels/Contents.html
@@ -33,7 +33,10 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
    &quot;The Fastest Pedestrian Detector in the West,&quot; BMVC 2010.
   [3] P. Dollár, R. Appel and W. Kienzle
    &quot;Crosstalk Cascades for Frame-Rate Pedestrian Detection,&quot; ECCV 2012.
+  [4] P. Dollár, R. Appel, S. Belongie and P. Perona
+   &quot;Fast Feature Pyramids for Object Detection&quot;, PAMI 2014.
  Please cite a subset of the above papers if you end up using the code.
+ The PAMI 2014 paper has the most thorough and up to date descriptions.
  Code written and maintained by Piotr Dollar and Ron Appel.
 
  Channels:

diff --git a/doc/channels/chnsPyramid.html b/doc/channels/chnsPyramid.html
@@ -46,8 +46,8 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
  can be used to approximate feature responses at nearby scales. The
  approximation is accurate at least within an entire scale octave. For
  details and to understand why this unexpected result holds, please see:
-   P. Dollár, S. Belongie and P. Perona
-   &quot;The Fastest Pedestrian Detector in the West,&quot; BMVC 2010.
+   P. Dollár, R. Appel, S. Belongie and P. Perona
+   &quot;Fast Feature Pyramids for Object Detection&quot;, PAMI 2014.
 
  The parameter &quot;nApprox&quot; determines how many intermediate scales are
  approximated using the techniques described in the above paper. Roughly
@@ -131,8 +131,8 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
 
  See also <a href="chnsCompute.html" class="code" title="function chns = chnsCompute( I, varargin )">chnsCompute</a>, <a href="chnsScaling.html" class="code" title="function [lambdas,as,scales,fs] = chnsScaling( pChns, Is, show )">chnsScaling</a>, <a href="convTri.html" class="code" title="function J = convTri( I, r, s, nomex )">convTri</a>, <a href="imPad.html" class="code" title="function J = imPad( I, pad, type )">imPad</a>
 
- Piotr's Image&amp;Video Toolbox      Version 3.02
- Copyright 2012 Piotr Dollar &amp; Ron Appel.  [pdollar-at-caltech.edu]
+ Piotr's Image&amp;Video Toolbox      Version 3.25
+ Copyright 2013 Piotr Dollar &amp; Ron Appel.  [pdollar-at-caltech.edu]
  Please email me if you find bugs, or have suggestions or questions!
  Licensed under the Simplified BSD License [see external/bsd.txt]</pre></div>
 

diff --git a/doc/channels/chnsScaling.html b/doc/channels/chnsScaling.html
@@ -30,8 +30,8 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
  can be used to approximate feature responses at nearby scales. The
  approximation is accurate at least within an entire scale octave. For
  details and to understand why this unexpected result holds, please see:
-   P. Dollár, S. Belongie and P. Perona
-   &quot;The Fastest Pedestrian Detector in the West,&quot; BMVC 2010.
+   P. Dollár, R. Appel, S. Belongie and P. Perona
+   &quot;Fast Feature Pyramids for Object Detection&quot;, PAMI 2014.
 
  This function computes channels at multiple image scales and plots the
  resulting power law scaling. The purpose of this function is two-fold:
@@ -62,14 +62,14 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
   fs             - [nImages x nScales x nTypes] array of feature means
 
  EXAMPLE
-  sDir='E:/code/detector/pedsData/train/negBig';
+  sDir = 'data/Inria/train/neg/';
   Is = fevalImages( @(x) {x}, {}, sDir, 'I', 'png', 0, 200 );
   p = chnsCompute(); lambdas = chnsScaling( p, Is, 1 );
 
  See also <a href="chnsCompute.html" class="code" title="function chns = chnsCompute( I, varargin )">chnsCompute</a>, <a href="chnsPyramid.html" class="code" title="function pyramid = chnsPyramid( I, varargin )">chnsPyramid</a>, fevalImages
 
- Piotr's Image&amp;Video Toolbox      Version 3.02
- Copyright 2012 Piotr Dollar &amp; Ron Appel.  [pdollar-at-caltech.edu]
+ Piotr's Image&amp;Video Toolbox      Version 3.25
+ Copyright 2013 Piotr Dollar &amp; Ron Appel.  [pdollar-at-caltech.edu]
  Please email me if you find bugs, or have suggestions or questions!
  Licensed under the Simplified BSD License [see external/bsd.txt]</pre></div>
 

diff --git a/doc/detector/Contents.html b/doc/detector/Contents.html
@@ -33,7 +33,10 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
    &quot;The Fastest Pedestrian Detector in the West,&quot; BMVC 2010.
   [3] P. Dollár, R. Appel and W. Kienzle
    &quot;Crosstalk Cascades for Frame-Rate Pedestrian Detection,&quot; ECCV 2012.
+  [4] P. Dollár, R. Appel, S. Belongie and P. Perona
+   &quot;Fast Feature Pyramids for Object Detection&quot;, PAMI 2014.
  Please cite a subset of the above papers if you end up using the code.
+ The PAMI 2014 paper has the most thorough and up to date descriptions.
  Code written and maintained by Piotr Dollar and Ron Appel.
 
  Aggregate channel features object detector:

diff --git a/doc/detector/acfTrain.html b/doc/detector/acfTrain.html
@@ -25,11 +25,11 @@ <h2><a name="_synopsis"></a>SYNOPSIS <a href="#_top"><img alt="^" border="0" src
 <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="0" src="../up.png"></a></h2>
 <div class="fragment"><pre class="comment"> Train aggregate channel features object detector.
 
- Train aggregate channel features (ACF) object detector:
-  P. Dollár, Z. Tu, P. Perona and S. Belongie
-  &quot;Integral Channel Features&quot;, BMVC 2009.
- The exact details of the algorithm below differ substantially from the
- BMVC09 paper, a more up-to-date method writeup is not yet available.
+ Train aggregate channel features (ACF) object detector as described in:
+  P. Dollár, R. Appel, S. Belongie and P. Perona
+   &quot;Fast Feature Pyramids for Object Detection&quot;, PAMI 2014.
+ The ACF detector is fast (30 fps on a single core) and achieves top
+ accuracy on rigid object detection.
 
  Takes a set of parameters opts (described in detail below) and trains a
  detector from start to finish including performing multiple rounds of
@@ -129,7 +129,7 @@ <h2><a name="_description"></a>DESCRIPTION <a href="#_top"><img alt="^" border="
  See also <a href="acfDetect.html" class="code" title="function bbs = acfDetect( I, detector, fileName )">acfDetect</a>, <a href="acfDemoInria.html" class="code" title="">acfDemoInria</a>, <a href="acfModify.html" class="code" title="function detector = acfModify( detector, varargin )">acfModify</a>, <a href="acfTest.html" class="code" title="function [miss,roc,gt,dt] = acfTest( varargin )">acfTest</a>, chnsCompute,
  chnsPyramid, adaBoostTrain, <a href="bbGt.html" class="code" title="function varargout = bbGt( action, varargin )">bbGt</a>, <a href="bbNms.html" class="code" title="function bbs = bbNms( bbs, varargin )">bbNms</a>, jitterImage
 
- Piotr's Image&amp;Video Toolbox      Version 3.22
+ Piotr's Image&amp;Video Toolbox      Version 3.25
  Copyright 2013 Piotr Dollar &amp; Ron Appel.  [pdollar-at-caltech.edu]
  Please email me if you find bugs, or have suggestions or questions!
  Licensed under the Simplified BSD License [see external/bsd.txt]</pre></div>

diff --git a/doc/history.txt b/doc/history.txt
@@ -1,3 +1,7 @@
+Version 3.25 (15-Dec-2013)
+ -added pointer to PAMI14 work: chnsScaling.m, acfTrain.m, chnsPyramid.m
+ -minor: toolboxCompile.m
+
 Version 3.24 (10-Oct-2013)
  -classify/private/*.cpp: minor fixes for very large datasets
  -classify/forest*.m: added new options for splitting/applying 

diff --git a/doc/overview.html b/doc/overview.html
@@ -40,12 +40,11 @@ <h2>DOWNLOAD<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
  <p><strong>For version history click <a href="history.txt">here.</a>
  This code is licensed under the <a href="http://en.wikipedia.org/wiki/BSD_licenses" target="blank">Simplified BSD License.</a></strong></p>
  <ul>
-  <li><a href="../piotr_toolbox.zip">Version 3.24</a> &nbsp; Oct. 10, 2013 &nbsp; [current stable version]
+  <li><a href="../piotr_toolbox.zip">Version 3.25</a> &nbsp; Dec. 15, 2013 &nbsp; [current stable version]
   <li><a href="../piotr_toolbox_V1.03.zip">Version 1.03</a> &nbsp; May. 03, 2006 &nbsp; [compatible with cuboids code]
   <li><a href="../">All Versions</a>
  </ul>
- </div>
-
+  <p>Requires Matlab 2011b or later and Matlab's Image Processing Toolbox.</p>
 </div>
 
 <h2>INSTALL<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
@@ -54,8 +53,8 @@ <h2>INSTALL<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
  <br/>&nbsp; &gt;&gt; addpath(genpath('c:\toolbox')); savepath;</p>
  <p>If needed, run the compile script for the mex files:
    <br/>&nbsp; &gt;&gt; toolboxCompile;
-   <br/>Note: 64 bit Windows/Linux binaries are already included.
- </p> 
+  <br/>Note: 64 bit Windows/Linux binaries are already included.
+ </p>
 </div>
 
 <h2>CITE<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
@@ -73,7 +72,7 @@ <h2>CONTACT<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
 <div class=box>
  If you have a question or a suggestion, or if you find a bug,
  please email me at<font style="BACKGROUND-COLOR: #f5f5f5" color=#0000ff>
- pdollar-at-caltech.edu</font></div>
+ pdollar-at-caltech.edu</font>
 </div>
 
 <!-- Start of Google Analytics Code -->

diff --git a/external/history.txt b/external/history.txt
@@ -1,4 +1,4 @@
-Version NEW
+Version 3.25 (15-Dec-2013)
  -added pointer to PAMI14 work: chnsScaling.m, acfTrain.m, chnsPyramid.m
  -minor: toolboxCompile.m
 

diff --git a/external/m2html/templates/frame-piotr/overview.html b/external/m2html/templates/frame-piotr/overview.html
@@ -40,7 +40,7 @@ <h2>DOWNLOAD<a href="#_top"><img alt=^ src="up.png" border=0 /></a></h2>
  <p><strong>For version history click <a href="history.txt">here.</a>
  This code is licensed under the <a href="http://en.wikipedia.org/wiki/BSD_licenses" target="blank">Simplified BSD License.</a></strong></p>
  <ul>
-  <li><a href="../piotr_toolbox.zip">Version 3.24</a> &nbsp; Oct. 10, 2013 &nbsp; [current stable version]
+  <li><a href="../piotr_toolbox.zip">Version 3.25</a> &nbsp; Dec. 15, 2013 &nbsp; [current stable version]
   <li><a href="../piotr_toolbox_V1.03.zip">Version 1.03</a> &nbsp; May. 03, 2006 &nbsp; [compatible with cuboids code]
   <li><a href="../">All Versions</a>
  </ul>