1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
|
const char *help = "\
GMM (c) Samy Bengio & Co 2001\n\
\n\
This program is used to do decode simple speech recognition \n";
#include <torch/LexiconInfo.h>
#include <torch/EMTrainer.h>
#include <torch/HMM.h>
#include <torch/CmdLine.h>
#include <torch/Random.h>
#include <torch/DiskHTKDataSet.h>
#include <torch/HTKDataSet.h>
#include <torch/PhoneModels.h>
#include <torch/LanguageModel.h>
#include <torch/LinearLexicon.h>
#include <torch/BeamSearchDecoder.h>
#include <torch/DiagonalGMM.h>
#include <torch/string_utils.h>
#include <torch/EditDistanceMeasurer.h>
#include <torch/WordSegMeasurer.h>
#include <torch/FileListCmdOption.h>
using namespace Torch;
#include "speech_include.cc"
int main(int argc, char **argv)
{
char* saved_model;
char* phoneme_name;
char* lex_name;
char* silence_phone;
char* sp_phone;
char* silence_word;
char* start_word;
char* end_word;
real log_word_entrance_penalty;
bool no_self_transitions;
int max_load;
char *dir_name;
bool htk_model;
bool disk;
bool print_timing;
char* lm_name;
Allocator *allocator = new Allocator;
FileListCmdOption input_file_list("file name", "the list of inputs files or one data file");
input_file_list.isArgument(true);
FileListCmdOption target_file_list("file name", "the list of target files or one target file");
target_file_list.isArgument(true);
//===============================================================
//=================== The command-line ==========================
//===============================================================
// Construct the command line
CmdLine cmd;
// Put the help line at the beginning
cmd.info(help);
// Train mode
cmd.addText("\nArguments:");
cmd.addSCmdArg("saved_model", &saved_model, "the saved model");
cmd.addSCmdArg("phoneme_name", &phoneme_name, "the list of phonemes file");
cmd.addSCmdArg("lex_name", &lex_name, "the lexicon file");
cmd.addCmdOption(&input_file_list);
cmd.addCmdOption(&target_file_list);
cmd.addText("\nModel Options:");
cmd.addSCmdOption("-lm", &lm_name,"", "optional language model");
cmd.addSCmdOption("-silence_phone", &silence_phone,"h#", "name of silence phone");
cmd.addSCmdOption("-sp_phone", &sp_phone,"", "name of short pause phoneme");
cmd.addSCmdOption("-silence_word", &silence_word,"sil", "name of silence word");
cmd.addSCmdOption("-start_word", &start_word,"<s>", "name of start word");
cmd.addSCmdOption("-end_word", &end_word,"</s>", "name of end word");
cmd.addRCmdOption("-log_word_entrance_penalty", &log_word_entrance_penalty, -15., "log of the word entrance penalty");
cmd.addBCmdOption("-no_self_transitions", &no_self_transitions, false, "do not admit grammar self transitions");
cmd.addText("\nMisc Options:");
cmd.addICmdOption("-load", &max_load, -1, "max number of examples to load for train");
cmd.addSCmdOption("-dir", &dir_name, ".", "directory to save measures");
cmd.addBCmdOption("-htk_model", &htk_model, false,"model file is in HTK format");
cmd.addBCmdOption("-disk", &disk, false, "keep data on disk");
cmd.addBCmdOption("-print_timing", &print_timing, false, "print timing decoding information");
// Read the command line
cmd.read(argc, argv);
cmd.setWorkingDirectory(dir_name);
DiskXFile::setBigEndianMode();
//====================================================================
//=================== Testing Mode =================================
//====================================================================
Random::seed();
// read lexicon
if (strlen(sp_phone) == 0) sp_phone = NULL;
if (strlen(silence_phone) == 0)
silence_phone = NULL;
LexiconInfo lexi(phoneme_name,silence_phone,sp_phone,lex_name,start_word,end_word);
int silence_index = lexi.vocabulary->getIndex(silence_word);
//====================================================================
//=================== Create the DataSet ... =========================
//====================================================================
// some basic tests on the files
if (input_file_list.n_files != target_file_list.n_files) {
error("the input and target files should have the same number of files (%d != %d)\n",input_file_list.n_files,target_file_list.n_files);
}
for (int i=0;i<input_file_list.n_files;i++) {
char* si = strRemoveSuffix(input_file_list.file_names[i]);
char* st = strRemoveSuffix(target_file_list.file_names[i]);
if (strcmp(strBaseName(si),strBaseName(st)))
warning("input file (%s) do not correspond to target file (%s)",input_file_list.file_names[i],target_file_list.file_names[i]);
free(si);
free(st);
}
DataSet* data;
int n_per_frame;
if (disk) {
data = (DataSet*)new(allocator) DiskHTKDataSet(input_file_list.file_names,target_file_list.file_names,input_file_list.n_files, true, max_load, &lexi);
n_per_frame = ((DiskHTKDataSet*)data)->n_per_frame;
} else {
data = (DataSet*)new(allocator) HTKDataSet(input_file_list.file_names,target_file_list.file_names, input_file_list.n_files,true, max_load, &lexi);
n_per_frame = ((HTKDataSet*)data)->n_per_frame;
}
//=================== Create the HMM... =========================
SpeechHMM* shmm = NULL;
if (htk_model)
shmm = newSpeechHMMFromHTK(saved_model,&lexi,NULL,0,allocator);
else
shmm = newSpeechHMMFromTorch(saved_model,&lexi,NULL,0,allocator,&cmd);
PhoneModels pm(shmm);
LinearLexicon lex(shmm,&pm);
LanguageModel *lm = NULL;
if ((strcmp(lm_name,"")))
lm = new LanguageModel(2,lexi.vocabulary,lm_name,1.0);
BeamSearchDecoder decoder(&lex,lm,log_word_entrance_penalty,LOG_ZERO,LOG_ZERO,false);
shmm->setDataSet(data);
shmm->eMIterInitialize();
EditDistance edit_d;
EditDistanceMeasurer edit_m(&edit_d,data,cmd.getXFile("decoder_edit"));
WordSegMeasurer word_m(NULL,data,cmd.getXFile("decoder_word"));
for (int i=0;i<data->n_examples;i++) {
data->setExample(i);
int num_result_words;
int* result_words;
int* result_words_times;
// remove some words from target sequence
int n_targets = data->targets->n_frames;
int* targets = (int*)allocator->alloc(sizeof(int)*n_targets);
int k=0;
for (int j=0;j<data->targets->n_frames;j++) {
int word = (int)data->targets->frames[j][0];
if (word != lexi.vocabulary->sent_start_index &&
word != silence_index &&
word != lexi.vocabulary->sent_end_index) {
targets[k++] = word;
}
}
n_targets = k;
edit_d.reset();
decoder.decode(data->inputs->frames,data->inputs->n_frames,&num_result_words,&result_words,&result_words_times);
// remove some words from obtained sequence
k=0;
for (int j=0;j<num_result_words;j++) {
int ti = result_words_times[j];
int word = result_words[j];
if (word != lexi.vocabulary->sent_start_index &&
word != silence_index &&
word != lexi.vocabulary->sent_end_index) {
result_words_times[k] = ti;
result_words[k++] = word;
}
}
num_result_words = k;
edit_d.distance(result_words,num_result_words,targets,n_targets);
edit_m.measureExample();
word_m.file->printf("obtained: ");
for (int j=0;j<num_result_words;j++)
word_m.file->printf("%s ",lexi.vocabulary->getWord(result_words[j]));
word_m.file->printf("\n");
if (word_m.print_targets) {
word_m.file->printf("desired: ");
for (int j=0;j<n_targets;j++)
word_m.file->printf("%s ",lexi.vocabulary->getWord(targets[j]));
word_m.file->printf("\n");
}
if (print_timing) {
int last = 0;
for (int j=0;j<num_result_words-1;j++) {
word_m.file->printf("%d %d %s\n",last*n_per_frame,result_words_times[j+1]*n_per_frame,lexi.vocabulary->getWord(result_words[j]));
last = result_words_times[j+1];
}
word_m.file->printf("%d %d %s\n",last*n_per_frame,data->inputs->n_frames*n_per_frame,lexi.vocabulary->getWord(result_words[num_result_words-1]));
}
word_m.file->flush();
/*
printf("obtained: ");
for (int j=0;j<num_result_words;j++) {
printf("%s ",lexi.vocabulary->getWord(result_words[j]));
}
printf("\n");
printf("desired: ");
for (int j=0;j<data->targets->n_frames;j++) {
printf("%s ",lexi.vocabulary->getWord((int)data->targets->frames[j][0]));
}
printf("\n");
*/
free(result_words);
free(result_words_times);
allocator->free(targets);
}
edit_m.measureIteration();
word_m.measureIteration();
edit_m.measureEnd();
word_m.measureEnd();
if (lm)
delete lm;
delete allocator;
return(0);
}
|
