1#!/bin/sh
2#####################################################-*-mode:shell-script-*-
3##                                                                       ##
4##                  Language Technologies Institute                      ##
5##                     Carnegie Mellon University                        ##
6##                         Copyright (c) 2002                            ##
7##                        All Rights Reserved.                           ##
8##                                                                       ##
9##  Permission is hereby granted, free of charge, to use and distribute  ##
10##  this software and its documentation without restriction, including   ##
11##  without limitation the rights to use, copy, modify, merge, publish,  ##
12##  distribute, sublicense, and/or sell copies of this work, and to      ##
13##  permit persons to whom this work is furnished to do so, subject to   ##
14##  the following conditions:                                            ##
15##   1. The code must retain the above copyright notice, this list of    ##
16##      conditions and the following disclaimer.                         ##
17##   2. Any modifications must be clearly marked as such.                ##
18##   3. Original authors' names are not deleted.                         ##
19##   4. The authors' names are not used to endorse or promote products   ##
20##      derived from this software without specific prior written        ##
21##      permission.                                                      ##
22##                                                                       ##
23##  CARNEGIE MELLON UNIVERSITY AND THE CONTRIBUTORS TO THIS WORK         ##
24##  DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING      ##
25##  ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT   ##
26##  SHALL CARNEGIE MELLON UNIVERSITY NOR THE CONTRIBUTORS BE LIABLE      ##
27##  FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES    ##
28##  WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN   ##
29##  AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION,          ##
30##  ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF       ##
31##  THIS SOFTWARE.                                                       ##
32##                                                                       ##
33###########################################################################
34##                                                                       ##
35##  Run SphinxTrain on data to build acoustic models then label it       ##
36##  with forced alignment                                                ##
37##                                                                       ##
38###########################################################################
39
40. etc/voice.defs
41
42if [ ! "$ESTDIR" ]
43then
44   echo "environment variable ESTDIR is unset"
45   echo "set it to your local speech tools directory e.g."
46   echo '   bash$ export ESTDIR=/home/awb/projects/speech_tools/'
47   echo or
48   echo '   csh% setenv ESTDIR /home/awb/projects/speech_tools/'
49   exit 1
50fi
51
52if [ ! "$SPHINXTRAINDIR" ]
53then
54   echo "environment variable SPHINXTRAINDIR is unset"
55   echo "set it to your local SphinxTrain src directory e.g."
56   echo '   bash$ export SPHINXTRAINDIR=/home/awb/projects/SphinxTrain/'
57   echo or
58   echo '   csh% setenv SPHINXTRAINDIR /home/awb/projects/SphinxTrain/'
59   exit 1
60fi
61
62if [ ! "$SPHINX2DIR" ]
63then
64   echo "environment variable SPHINX2DIR is unset"
65   echo "set it to your local Sphinx2 is installed e.g."
66   echo '   bash$ export SPHINX2DIR=/home/awb/projects/sphinx2/'
67   echo or
68   echo '   csh% setenv SPHINX2DIR /home/awb/projects/sphinx2/'
69   echo 'such that $SPHINX2DIR/bin/sphinx2-batch exists'
70   exit 1
71fi
72
73if [ $# = 0 ]
74then
75   $0 setup
76   $0 files
77   $0 feats
78   $0 train
79   $0 align
80   $0 labs
81   exit
82fi
83
84if [ "$1" = "setup" ]
85then
86
87  mkdir st
88  cd st
89
90  $SPHINXTRAINDIR/scripts_pl/setup_SphinxTrain $FV_VOICENAME
91  mkdir lab
92
93  cd ..
94fi
95
96#  dic, fileids, phone, filler, transcription
97if [ "$1" = "files" ]
98then
99   festival -b festvox/build_st.scm '(st_setup "etc/txt.done.data" "'$FV_VOICENAME'")'
100fi
101
102# Wave 2 feats
103if [ "$1" = "feats" ]
104then
105   for i in wav/*.wav
106   do
107      $ESTDIR/bin/ch_wave -otype nist -o st/$i $i
108   done
109   (cd st; ./bin/make_feats etc/$FV_VOICENAME.fileids)
110fi
111
112if [ "$1" = "train" ]
113then
114   cd st
115   ./scripts_pl/00.verify/verify_all.pl
116   ./scripts_pl/01.vector_quantize/slave.VQ.pl
117   ./scripts_pl/02.ci_schmm/slave_convg.pl
118   ./scripts_pl/03.makeuntiedmdef/make_untied_mdef.pl
119   ./scripts_pl/04.cd_schmm_untied/slave_convg.pl
120   ./scripts_pl/05.buildtrees/make_questions.pl
121   ./scripts_pl/05.buildtrees/slave.treebuilder.pl
122   ./scripts_pl/06.prunetree/slave.state-tie-er.pl
123   ./scripts_pl/07.cd-schmm/slave_convg.pl
124   ./scripts_pl/08.deleted-interpolation/deleted_interpolation.pl
125   ./scripts_pl/09.make_s2_models/make_s2_models.pl
126   cd ..
127fi
128
129# Alignment
130if [ "$1" = "align" ]
131then
132   echo "*align_all*" >st/etc/$FV_VOICENAME.align
133   cat st/etc/$FV_VOICENAME.transcription |
134   sed 's/<sil>//g' |
135   awk '{for (i=2; i<NF-2; i++)
136           printf("%s ",$i);
137         printf("%s\n",$(NF-2))}' >>st/etc/$FV_VOICENAME.align
138   echo "<sil> SIL" >st/etc/$FV_VOICENAME.sil
139   ./bin/sphinx_lab
140fi
141
142# Copy back lab files
143if [ "$1" = "labs" ]
144then
145   silence=`head -1 etc/mysilence`
146   for i in st/lab/*.lab
147   do
148     cat $i | sed 's/;.*$//;s/(.*$//' |
149              sed 's/SIL/'$silence'/;s/CAP//' |
150     awk '{if (NF == 3)
151             printf("%1.5f %s %s\n",0.012+$1,$2,$3);
152           else
153             print $0}' >lab/`basename $i`
154   done
155fi
156