trunk: Documentation changes including improvements to the look of the website; minor...

git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@4936 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8

diff --git a/egs/wsj/s5/steps/decode_fmllr.sh b/egs/wsj/s5/steps/decode_fmllr.sh
index 694203d3859fde58a40d4efcc0e8d02bdc22ce1d..97d32e6c59ba0808268badffb326346f8c51b9a9 100755 (executable)
--- a/egs/wsj/s5/steps/decode_fmllr.sh
+++ b/egs/wsj/s5/steps/decode_fmllr.sh
@@ -45,7 +45,9 @@ num_threads=1 # if >1, will use gmm-latgen-faster-parallel
 parallel_opts=  # If you supply num-threads, you should supply this too.
 skip_scoring=false
 scoring_opts=
-# End configuration section
+max_fmllr_jobs=25  # I've seen the fMLLR jobs overload NFS badly if the decoding
+                   # was started with a lot of many jobs, so we limit the number of
+                   # parallel jobs to 25 by default.  End configuration section
 echo "$0 $@"  # Print the command line for logging
 
 [ -f ./path.sh ] && . ./path.sh; # source the path.
@@ -147,7 +149,7 @@ esac
 ## Now get the first-pass fMLLR transforms.
 if [ $stage -le 1 ]; then
   echo "$0: getting first-pass fMLLR transforms."
-  $cmd JOB=1:$nj $dir/log/fmllr_pass1.JOB.log \
+  $cmd --max-jobs-run $max_fmllr_jobs JOB=1:$nj $dir/log/fmllr_pass1.JOB.log \
     gunzip -c $si_dir/lat.JOB.gz \| \
     lattice-to-post --acoustic-scale=$acwt ark:- ark:- \| \
     weight-silence-post $silence_weight $silphonelist $alignment_model ark:- ark:- \| \
@@ -183,7 +185,7 @@ fi
 ## $dir/trans.1, etc.
 if [ $stage -le 3 ]; then
   echo "$0: estimating fMLLR transforms a second time."
-  $cmd JOB=1:$nj $dir/log/fmllr_pass2.JOB.log \
+  $cmd --max-jobs-run $max_fmllr_jobs JOB=1:$nj $dir/log/fmllr_pass2.JOB.log \
     lattice-determinize-pruned$thread_string --acoustic-scale=$acwt --beam=4.0 \
     "ark:gunzip -c $dir/lat.tmp.JOB.gz|" ark:- \| \
     lattice-to-post --acoustic-scale=$acwt ark:- ark:- \| \

diff --git a/misc/logo/KaldiIco.png b/misc/logo/KaldiIco.png
new file mode 100644 (file)
index 0000000..4618361
Binary files /dev/null and b/misc/logo/KaldiIco.png differ

diff --git a/misc/logo/KaldiTextAndLogoSmall.png b/misc/logo/KaldiTextAndLogoSmall.png
new file mode 100644 (file)
index 0000000..4293180
Binary files /dev/null and b/misc/logo/KaldiTextAndLogoSmall.png differ

diff --git a/src/Doxyfile b/src/Doxyfile
index cfa6cbbc446671c4075528ffe41498db7a90f710..86d63b9e22966268d7d09461e81c5c02e78aaf41 100644 (file)
--- a/src/Doxyfile
+++ b/src/Doxyfile
@@ -17,7 +17,7 @@
 # The PROJECT_NAME tag is a single word (or a sequence of words surrounded 
 # by quotes) that should identify the project.
 
-PROJECT_NAME           = 'KALDI'
+PROJECT_NAME           = Kaldi
 
 # The PROJECT_NUMBER tag can be used to enter a project or revision number. 
 # This could be handy for archiving the generated documentation or 
@@ -25,6 +25,20 @@ PROJECT_NAME           = 'KALDI'
 
 PROJECT_NUMBER         = 
 
+
+# Using the PROJECT_BRIEF tag one can provide an optional one line description
+# for a project that appears at the top of each page and should give viewer
+# a quick idea about the purpose of the project. Keep the description short.
+
+# PROJECT_BRIEF          =  "Open source speech recognition"
+
+# With the PROJECT_LOGO tag one can specify an logo or icon that is
+# included in the documentation. The maximum height of the logo should not
+# exceed 55 pixels and the maximum width should not exceed 200 pixels.
+# Doxygen will copy the logo to the output directory.
+ 
+PROJECT_LOGO           =  ../misc/logo/KaldiTextAndLogoSmall.png
+
 # The OUTPUT_DIRECTORY tag is used to specify the (relative or absolute) 
 # base path where the generated documentation will be put. 
 # If a relative path is entered, it will be relative to the location 
@@ -639,7 +653,40 @@ HTML_FILE_EXTENSION    = .html
 # each generated HTML page. If it is left blank doxygen will generate a 
 # standard header.
 
-HTML_HEADER            = 
+## Note: doc/header.html is a modified version of the standard header which was generated
+## using "doxygen -w html header.html footer.html stylesheet.css Doxyfile" (note, this may
+## have to be updated if the doxygen version changes), and then the following line was added:
+## <link rel="icon" type="image/png" href="http://kaldi.sf.net/favicon.ico">
+
+HTML_HEADER            = doc/header.html
+
+
+# The HTML_COLORSTYLE_HUE tag controls the color of the HTML output.
+# Doxygen will adjust the colors in the style sheet and background images
+# according to this color. Hue is specified as an angle on a colorwheel,
+# see http://en.wikipedia.org/wiki/Hue for more information.
+# For instance the value 0 represents red, 60 is yellow, 120 is green,
+# 180 is cyan, 240 is blue, 300 purple, and 360 is red again.
+# The allowed range is 0 to 359.
+
+HTML_COLORSTYLE_HUE    = 26
+
+# The HTML_COLORSTYLE_SAT tag controls the purity (or saturation) of
+# the colors in the HTML output. For a value of 0 the output will use
+# grayscales only. A value of 255 will produce the most vivid colors.
+
+HTML_COLORSTYLE_SAT    = 80
+
+# The HTML_COLORSTYLE_GAMMA tag controls the gamma correction applied to
+# the luminance component of the colors in the HTML output. Values below
+# 100 gradually make the output lighter, whereas values above 100 make
+# the output darker. The value divided by 100 is the actual gamma applied,
+# so 80 represents a gamma of 0.8, The value 220 represents a gamma of 2.2,
+# and 100 does not change the gamma.
+
+HTML_COLORSTYLE_GAMMA  = 90
+
+
 
 # The HTML_FOOTER tag can be used to specify a personal HTML footer for 
 # each generated HTML page. If it is left blank doxygen will generate a 
@@ -654,8 +701,7 @@ HTML_FOOTER            =
 # the style sheet file to the HTML output directory, so don't put your own 
 # stylesheet in the HTML output directory as well, or it will be erased!
 
-HTML_STYLESHEET        = 
-
+HTML_STYLESHEET        =
 
 # If the GENERATE_HTMLHELP tag is set to YES, additional index files 
 # will be generated that can be used as input for tools like the 

diff --git a/src/doc/README b/src/doc/README
index e79f7cc86e25509cbd7a89c69c7bb65a0795df9d..4d145f9829a1375d0da2f0b4642dc7bad9d3ba2c 100644 (file)
--- a/src/doc/README
+++ b/src/doc/README
@@ -24,6 +24,7 @@ doxygen
 cp doc/*.pptx html/;
 
 if [[ $(hostname -f) == *.clsp.jhu.edu ]]; then
+   cp ../misc/logo/KaldiIco.png html/favicon.ico
    tar -czf html.tar.gz html
    scp html.tar.gz danielpovey@web.sourceforge.net:/home/project-web/kaldi/htdocs/
 
@@ -49,3 +50,10 @@ fi
 # I added figures that I manually excerpted from https://sourceforge.net/p/kaldi/mailman/kaldi-users/?viewmonth=201203
 # and https://sourceforge.net/p/kaldi/mailman/kaldi-developers/?viewmonth=201203
 # this is june 13, 2014, 6:11pm, check my email.
+
+# Note (RE adding favicon): I generated the default header files like this (from
+# src/) doxygen -w html header.html footer.html stylesheet.css Doxyfile then
+# moved the header.html to doc/ and edited it to include the following snippet,
+# and added it to the repo.
+#<link rel="icon" type="image/png" href="http://kaldi.sf.net/favicon.ico">
+

diff --git a/src/doc/build_setup.dox b/src/doc/build_setup.dox
index 42f48fba6315143ee8ee385d57470afd0c1d6aaa..86dca5cad693f40f734b75e219253bbffa68dfd4 100644 (file)
--- a/src/doc/build_setup.dox
+++ b/src/doc/build_setup.dox
@@ -144,7 +144,7 @@ preprocessor variables, setting compile options, linking with libraries, and so
 \section build_setup_platforms Which platforms has Kaldi been compiled on?
 
 We have compiled Kaldi on Windows, Cygwin, various flavors of Linux (including
-Ubuntu, CentOS, Debian and SUSE), and Darwin. We recommend you use g++ version
-4.4 or above for the source to compile.
+Ubuntu, CentOS, Debian, Red Hat and SUSE), and Darwin. We recommend you use g++ version
+4.4 or above, although other compilers such as llvm and Intel's icc are also known to work.
 
 */

diff --git a/src/doc/header.html b/src/doc/header.html
new file mode 100644 (file)
index 0000000..9e496f2
--- /dev/null
+++ b/src/doc/header.html
@@ -0,0 +1,51 @@
+<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
+<html xmlns="http://www.w3.org/1999/xhtml">
+<head>
+<meta http-equiv="Content-Type" content="text/xhtml;charset=UTF-8"/>
+<meta http-equiv="X-UA-Compatible" content="IE=9"/>
+<!--BEGIN PROJECT_NAME--><title>$projectname: $title</title><!--END PROJECT_NAME-->
+<!--BEGIN !PROJECT_NAME--><title>$title</title><!--END !PROJECT_NAME-->
+<link href="$relpath$tabs.css" rel="stylesheet" type="text/css"/>
+<link rel="icon" href="favicon.ico" type="image/x-icon" />
+<script type="text/javascript" src="$relpath$jquery.js"></script>
+<script type="text/javascript" src="$relpath$dynsections.js"></script>
+$treeview
+$search
+$mathjax
+<link href="stylesheet.css" rel="stylesheet" type="text/css" /> 
+</head>
+<body>
+<div id="top"><!-- do not remove this div, it is closed by doxygen! -->
+
+<!--BEGIN TITLEAREA-->
+<div id="titlearea">
+<table cellspacing="0" cellpadding="0">
+ <tbody>
+ <tr style="height: 56px;">
+  <!--BEGIN PROJECT_LOGO-->
+  <td id="projectlogo"><img alt="Logo" src="$relpath$$projectlogo"/ style="padding: 4px 5px 1px 5px"></td>
+  <!--END PROJECT_LOGO-->
+  <!--BEGIN PROJECT_NAME-->
+  <td style="padding-left: 0.5em;">
+   <div id="projectname" style="display:none">$projectname
+   <!--BEGIN PROJECT_NUMBER-->&#160;<span id="projectnumber">$projectnumber</span><!--END PROJECT_NUMBER-->
+   </div>
+   <!--BEGIN PROJECT_BRIEF--><div id="projectbrief">$projectbrief</div><!--END PROJECT_BRIEF-->
+  </td>
+  <!--END PROJECT_NAME-->
+    <td style="padding-left: 0.5em;">
+    <div id="projectbrief" style="display:none">$projectbrief</div>
+    </td>
+   <!--END PROJECT_BRIEF-->
+  <!--END !PROJECT_NAME-->
+  <!--BEGIN DISABLE_INDEX-->
+   <!--BEGIN SEARCHENGINE-->
+   <td>$searchbox</td>
+   <!--END SEARCHENGINE-->
+  <!--END DISABLE_INDEX-->
+ </tr>
+ </tbody>
+</table>
+</div>
+<!--END TITLEAREA-->
+<!-- end header part -->

diff --git a/src/doc/mainpage.dox b/src/doc/mainpage.dox
index 472ad00ccc2bcf0861b28b53846d93c790fee95c..2f9bc3b273e910148b90b3ce910ef7eef2416a23 100644 (file)
--- a/src/doc/mainpage.dox
+++ b/src/doc/mainpage.dox
@@ -30,10 +30,7 @@
 /**
   \mainpage Kaldi
     
-   Please see the \ref install_warning "instructions" on upgrading your repository to the
-   new location, following our upgrade to the "new" Sourceforge.
-
-   (see also Kaldi's <a href=http://sourceforge.net/projects/kaldi/> project page on Sourceforge </a>,
+   See also Kaldi's <a href=http://sourceforge.net/projects/kaldi/>project page on Sourceforge</a>,
     and <a href=http://kaldi-asr.org>kaldi-asr.org</a>  where you can download pre-built models.
 
   <p>
@@ -42,7 +39,6 @@
    - \subpage install 
    - \subpage dependencies
    - \subpage legal
-   - \subpage roadmap
    - \subpage tutorial
    - \subpage data_prep
    - \subpage build_setup 

diff --git a/src/doc/roadmap.dox b/src/doc/roadmap.dox
deleted file mode 100644 (file)
index d93fbfa..0000000
--- a/src/doc/roadmap.dox
+++ /dev/null
@@ -1,121 +0,0 @@
-// doc/roadmap.dox
-
-// Copyright 2014  Johns Hopkins University (author: Daniel Povey)
-
-// See ../../COPYING for clarification regarding multiple authors
-//
-// Licensed under the Apache License, Version 2.0 (the "License");
-// you may not use this file except in compliance with the License.
-// You may obtain a copy of the License at
-
-//  http://www.apache.org/licenses/LICENSE-2.0
-
-// THIS CODE IS PROVIDED *AS IS* BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
-// KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY IMPLIED
-// WARRANTIES OR CONDITIONS OF TITLE, FITNESS FOR A PARTICULAR PURPOSE,
-// MERCHANTABLITY OR NON-INFRINGEMENT.
-// See the Apache 2 License for the specific language governing permissions and
-// limitations under the License.
-
-
-/**
-   \page roadmap  Plans for Kaldi development
-
-  This page describes the features we are currently working on or have recently
-  completed, and how we aim to develop Kaldi in the future.
-
- \section roadmap_current  Features we are currently working on
-
-   \subsection roadmap_current_online Online decoding
-
-  By online decoding, we mean decoding that can take in audio data frame by frame and
-  output a result with minimal latency.  Currently, in ^/sandbox/online (see src/online2
-  and egs/rm/s5/local/run_online_decoding.sh), we are working on an improved framework
-  for online decoding.  This is already functional in terms of GMM-based decoding.
-  The things that are currently preventing the setup from being merged back to trunk are:
-
-     - Issues with online pitch feature generation.  We need to verify that the modified
-       pitch-feature extraction code does not cause a degradation for our existing offline 
-       recipes
-     - We need to finalize the neural net version of the example scripts.  Dan was hoping to
-       have finished this by now, but got involved with something else (speeding up the
-       neural network training).  Also, he is undecided on whether to first work on an
-       improved version of the neural network recipe that works well without fMLLR adaptation
-       (e.g. using online-estimated iVectors)
-
-   It's currently not clear how long it will take to merge it back to trunk; it
-  could be a couple of weeks to a couple of months, but anyone is free to use
-  the sandbox/online version and they should be confident that this will
-  eventually be merged back to trunk.
-
-  \subsection roadmap_current_nnet Neural network related changes
-
-  There are various things that are being worked on by both Dan and Karel but are not
-  likely to be finalized soon.  Some of this work relates to convolutional networks.
-
-  Dan has recently (June 2014) improved the speed of his neural network training
-  setup (see \ref dnn2) by improving the way the preconditioning was applied.
-  Also relating to Dan's neural setup, Samuel Zhang recently added an option
-  --first-component-power to train_pnorm_fast.sh; setting this to 0.5 seems to
-  give improvement if there is a reasonable amount of data (this takes the
-  output of the first p-norm component and raises it the the power 0.5).
-
-
-  \subsection roadmap_current_server  Accessing Kaldi over the internet
-
-  A couple of people are working on ways to access Kaldi over the Internet
-  (e.g. a REST API).  Although it may not make sense to make these part of
-  Kaldi's repository, we hope to at least point to an external project that
-  makes it easy to do this.  Note, we're not talking about setting up a public
-  server at this point, just showing others how to do so.
-  We're also interested in more telephone-oriented protocols such as MRCP, and
-  if anyone wants to work on that, we would welcome it.
-
- \section roadmap_like Features we would like to work on, if we had the time
-
-   \subsection roadmap_like_faster Faster decoders
-
-  An outstanding issue is, we'd like the decoding to be faster.  The question is
-  how to do this in a way that's not too specific to one type of model.  One way
-  this could be done is to convert the OpenFst decoding graph into some other,
-  more optimized data-structure.
-
-   \subsection roadmap_like_examples More example scripts
-
-  We are constantly adding new example scripts for new databases, and many of these are added by
-  new contributors to the Kaldi project.  We welcome such contributions.
-
-
-
- \section roadmap_wont Features we do not currently plan to work on
-
-   \subsection roadmap_wont_scripting A scripting layer (Python etc.).
-
-    Our approach is that if you need something that's not supported in the
-    current command-line programs, you just add a new command-line program.
-    This works well for us.  Maintaining a parallel set of, say, Python-wrapped
-    example scripts would be too much work.  However, people have created
-    examples of calling Kaldi code from other languages (Python, Perl, Java) for
-    various specific purposes.  Ask us for details.
-
-  \subsection roadmap_wont_nlp  Natural language processing, language model estimation, etc.
-
-   We don't think it makes sense for Kaldi to try to be everything, and NLP and
-   language model estimation are two examples of things that we probably don't
-   plan to do in the near future.  (However, new facilities for using externally
-   generated language models are definitely on the table).  Kaldi is primarily a
-   speech recognition project, and we plan to keep it that way, although we did
-   create code and example scripts for speaker identification and language
-   identification (see egs/sre08/ and egs/lre/), since these technologies have a
-   lot in common with speech recognition.  We are also dabbling in some
-   computer-vision (there is an MNIST example in ^/sandbox/convnets):
-   for now this is mostly just a convenient way to test our ideas related to
-   convolutional networks on vision tasks, rather than a major pivot towards
-   computer vision.
-
-
-   
-  
-
-
-*/