online2-wav-nnet3-latgen-grammar.cc
Go to the documentation of this file.
1 // online2bin/online2-wav-nnet3-latgen-grammar.cc
2 
3 // Copyright 2014-2018 Johns Hopkins University (author: Daniel Povey)
4 // 2016 Api.ai (Author: Ilya Platonov)
5 
6 // See ../../COPYING for clarification regarding multiple authors
7 //
8 // Licensed under the Apache License, Version 2.0 (the "License");
9 // you may not use this file except in compliance with the License.
10 // You may obtain a copy of the License at
11 //
12 // http://www.apache.org/licenses/LICENSE-2.0
13 //
14 // THIS CODE IS PROVIDED *AS IS* BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
15 // KIND, EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION ANY IMPLIED
16 // WARRANTIES OR CONDITIONS OF TITLE, FITNESS FOR A PARTICULAR PURPOSE,
17 // MERCHANTABLITY OR NON-INFRINGEMENT.
18 // See the Apache 2 License for the specific language governing permissions and
19 // limitations under the License.
20 
21 #include "feat/wave-reader.h"
24 #include "online2/onlinebin-util.h"
25 #include "online2/online-timing.h"
27 #include "fstext/fstext-lib.h"
28 #include "lat/lattice-functions.h"
29 #include "util/kaldi-thread.h"
30 #include "nnet3/nnet-utils.h"
31 #include "decoder/grammar-fst.h"
32 
33 namespace kaldi {
34 
35 void GetDiagnosticsAndPrintOutput(const std::string &utt,
36  const fst::SymbolTable *word_syms,
37  const CompactLattice &clat,
38  int64 *tot_num_frames,
39  double *tot_like) {
40  if (clat.NumStates() == 0) {
41  KALDI_WARN << "Empty lattice.";
42  return;
43  }
44  CompactLattice best_path_clat;
45  CompactLatticeShortestPath(clat, &best_path_clat);
46 
47  Lattice best_path_lat;
48  ConvertLattice(best_path_clat, &best_path_lat);
49 
50  double likelihood;
51  LatticeWeight weight;
52  int32 num_frames;
53  std::vector<int32> alignment;
54  std::vector<int32> words;
55  GetLinearSymbolSequence(best_path_lat, &alignment, &words, &weight);
56  num_frames = alignment.size();
57  likelihood = -(weight.Value1() + weight.Value2());
58  *tot_num_frames += num_frames;
59  *tot_like += likelihood;
60  KALDI_VLOG(2) << "Likelihood per frame for utterance " << utt << " is "
61  << (likelihood / num_frames) << " over " << num_frames
62  << " frames.";
63 
64  if (word_syms != NULL) {
65  std::cerr << utt << ' ';
66  for (size_t i = 0; i < words.size(); i++) {
67  std::string s = word_syms->Find(words[i]);
68  if (s == "")
69  KALDI_ERR << "Word-id " << words[i] << " not in symbol table.";
70  std::cerr << s << ' ';
71  }
72  std::cerr << std::endl;
73  }
74 }
75 
76 }
77 
78 int main(int argc, char *argv[]) {
79  try {
80  using namespace kaldi;
81  using namespace fst;
82 
83  typedef kaldi::int32 int32;
84  typedef kaldi::int64 int64;
85 
86  const char *usage =
87  "Reads in wav file(s) and simulates online decoding with neural nets\n"
88  "(nnet3 setup), with optional iVector-based speaker adaptation and\n"
89  "optional endpointing. Note: some configuration values and inputs are\n"
90  "set via config files whose filenames are passed as options.\n"
91  "This program like online2-wav-nnet3-latgen-faster but when the FST to\n"
92  "be decoded is of type GrammarFst.\n"
93  "\n"
94  "Usage: online2-wav-nnet3-latgen-grammar [options] <nnet3-in> <fst-in> "
95  "<spk2utt-rspecifier> <wav-rspecifier> <lattice-wspecifier>\n"
96  "The spk2utt-rspecifier can just be <utterance-id> <utterance-id> if\n"
97  "you want to decode utterance by utterance.\n";
98 
99  ParseOptions po(usage);
100 
101  std::string word_syms_rxfilename;
102 
103  // feature_opts includes configuration for the iVector adaptation,
104  // as well as the basic features.
107  LatticeFasterDecoderConfig decoder_opts;
108  OnlineEndpointConfig endpoint_opts;
109 
110  BaseFloat chunk_length_secs = 0.18;
111  bool do_endpointing = false;
112  bool online = true;
113 
114  po.Register("chunk-length", &chunk_length_secs,
115  "Length of chunk size in seconds, that we process. Set to <= 0 "
116  "to use all input in one chunk.");
117  po.Register("word-symbol-table", &word_syms_rxfilename,
118  "Symbol table for words [for debug output]");
119  po.Register("do-endpointing", &do_endpointing,
120  "If true, apply endpoint detection");
121  po.Register("online", &online,
122  "You can set this to false to disable online iVector estimation "
123  "and have all the data for each utterance used, even at "
124  "utterance start. This is useful where you just want the best "
125  "results and don't care about online operation. Setting this to "
126  "false has the same effect as setting "
127  "--use-most-recent-ivector=true and --greedy-ivector-extractor=true "
128  "in the file given to --ivector-extraction-config, and "
129  "--chunk-length=-1.");
130  po.Register("num-threads-startup", &g_num_threads,
131  "Number of threads used when initializing iVector extractor.");
132 
133  feature_opts.Register(&po);
134  decodable_opts.Register(&po);
135  decoder_opts.Register(&po);
136  endpoint_opts.Register(&po);
137 
138 
139  po.Read(argc, argv);
140 
141  if (po.NumArgs() != 5) {
142  po.PrintUsage();
143  return 1;
144  }
145 
146  std::string nnet3_rxfilename = po.GetArg(1),
147  fst_rxfilename = po.GetArg(2),
148  spk2utt_rspecifier = po.GetArg(3),
149  wav_rspecifier = po.GetArg(4),
150  clat_wspecifier = po.GetArg(5);
151 
152  OnlineNnet2FeaturePipelineInfo feature_info(feature_opts);
153  if (!online) {
156  chunk_length_secs = -1.0;
157  }
158 
159  Matrix<double> global_cmvn_stats;
160  if (feature_info.global_cmvn_stats_rxfilename != "")
162  &global_cmvn_stats);
163 
164  TransitionModel trans_model;
165  nnet3::AmNnetSimple am_nnet;
166  {
167  bool binary;
168  Input ki(nnet3_rxfilename, &binary);
169  trans_model.Read(ki.Stream(), binary);
170  am_nnet.Read(ki.Stream(), binary);
171  SetBatchnormTestMode(true, &(am_nnet.GetNnet()));
172  SetDropoutTestMode(true, &(am_nnet.GetNnet()));
174  }
175 
176  // this object contains precomputed stuff that is used by all decodable
177  // objects. It takes a pointer to am_nnet because if it has iVectors it has
178  // to modify the nnet to accept iVectors at intervals.
179  nnet3::DecodableNnetSimpleLoopedInfo decodable_info(decodable_opts,
180  &am_nnet);
181 
182 
184  ReadKaldiObject(fst_rxfilename, &fst);
185 
186  fst::SymbolTable *word_syms = NULL;
187  if (word_syms_rxfilename != "")
188  if (!(word_syms = fst::SymbolTable::ReadText(word_syms_rxfilename)))
189  KALDI_ERR << "Could not read symbol table from file "
190  << word_syms_rxfilename;
191 
192  int32 num_done = 0, num_err = 0;
193  double tot_like = 0.0;
194  int64 num_frames = 0;
195 
196  SequentialTokenVectorReader spk2utt_reader(spk2utt_rspecifier);
197  RandomAccessTableReader<WaveHolder> wav_reader(wav_rspecifier);
198  CompactLatticeWriter clat_writer(clat_wspecifier);
199 
200  OnlineTimingStats timing_stats;
201 
202  for (; !spk2utt_reader.Done(); spk2utt_reader.Next()) {
203  std::string spk = spk2utt_reader.Key();
204  const std::vector<std::string> &uttlist = spk2utt_reader.Value();
205 
206  OnlineIvectorExtractorAdaptationState adaptation_state(
207  feature_info.ivector_extractor_info);
208  OnlineCmvnState cmvn_state(global_cmvn_stats);
209 
210  for (size_t i = 0; i < uttlist.size(); i++) {
211  std::string utt = uttlist[i];
212  if (!wav_reader.HasKey(utt)) {
213  KALDI_WARN << "Did not find audio for utterance " << utt;
214  num_err++;
215  continue;
216  }
217  const WaveData &wave_data = wav_reader.Value(utt);
218  // get the data for channel zero (if the signal is not mono, we only
219  // take the first channel).
220  SubVector<BaseFloat> data(wave_data.Data(), 0);
221 
222  OnlineNnet2FeaturePipeline feature_pipeline(feature_info);
223  feature_pipeline.SetAdaptationState(adaptation_state);
224  feature_pipeline.SetCmvnState(cmvn_state);
225 
226  OnlineSilenceWeighting silence_weighting(
227  trans_model,
228  feature_info.silence_weighting_config,
229  decodable_opts.frame_subsampling_factor);
230 
232  decoder_opts, trans_model,
233  decodable_info, fst, &feature_pipeline);
234 
235  OnlineTimer decoding_timer(utt);
236 
237  BaseFloat samp_freq = wave_data.SampFreq();
238  int32 chunk_length;
239  if (chunk_length_secs > 0) {
240  chunk_length = int32(samp_freq * chunk_length_secs);
241  if (chunk_length == 0) chunk_length = 1;
242  } else {
243  chunk_length = std::numeric_limits<int32>::max();
244  }
245 
246  int32 samp_offset = 0;
247  std::vector<std::pair<int32, BaseFloat> > delta_weights;
248 
249  while (samp_offset < data.Dim()) {
250  int32 samp_remaining = data.Dim() - samp_offset;
251  int32 num_samp = chunk_length < samp_remaining ? chunk_length
252  : samp_remaining;
253 
254  SubVector<BaseFloat> wave_part(data, samp_offset, num_samp);
255  feature_pipeline.AcceptWaveform(samp_freq, wave_part);
256 
257  samp_offset += num_samp;
258  decoding_timer.WaitUntil(samp_offset / samp_freq);
259  if (samp_offset == data.Dim()) {
260  // no more input. flush out last frames
261  feature_pipeline.InputFinished();
262  }
263 
264  if (silence_weighting.Active() &&
265  feature_pipeline.IvectorFeature() != NULL) {
266  silence_weighting.ComputeCurrentTraceback(decoder.Decoder());
267  silence_weighting.GetDeltaWeights(feature_pipeline.NumFramesReady(),
268  &delta_weights);
269  feature_pipeline.IvectorFeature()->UpdateFrameWeights(delta_weights);
270  }
271 
272  decoder.AdvanceDecoding();
273 
274  if (do_endpointing && decoder.EndpointDetected(endpoint_opts)) {
275  break;
276  }
277  }
278  decoder.FinalizeDecoding();
279 
280  CompactLattice clat;
281  bool end_of_utterance = true;
282  decoder.GetLattice(end_of_utterance, &clat);
283 
284  GetDiagnosticsAndPrintOutput(utt, word_syms, clat,
285  &num_frames, &tot_like);
286 
287  decoding_timer.OutputStats(&timing_stats);
288 
289  // In an application you might avoid updating the adaptation state if
290  // you felt the utterance had low confidence. See lat/confidence.h
291  feature_pipeline.GetAdaptationState(&adaptation_state);
292  feature_pipeline.GetCmvnState(&cmvn_state);
293 
294  // we want to output the lattice with un-scaled acoustics.
295  BaseFloat inv_acoustic_scale =
296  1.0 / decodable_opts.acoustic_scale;
297  ScaleLattice(AcousticLatticeScale(inv_acoustic_scale), &clat);
298 
299  clat_writer.Write(utt, clat);
300  KALDI_LOG << "Decoded utterance " << utt;
301  num_done++;
302  }
303  }
304  timing_stats.Print(online);
305 
306  KALDI_LOG << "Decoded " << num_done << " utterances, "
307  << num_err << " with errors.";
308  KALDI_LOG << "Overall likelihood per frame was " << (tot_like / num_frames)
309  << " per frame over " << num_frames << " frames.";
310  delete word_syms; // will delete if non-NULL.
311  return (num_done != 0 ? 0 : 1);
312  } catch(const std::exception& e) {
313  std::cerr << e.what();
314  return -1;
315  }
316 } // main()
int32 words[kMaxOrder]
This code computes Goodness of Pronunciation (GOP) and extracts phone-level pronunciation feature for...
Definition: chain.dox:20
void CollapseModel(const CollapseModelConfig &config, Nnet *nnet)
This function modifies the neural net for efficiency, in a way that suitable to be done in test time...
Definition: nnet-utils.cc:2100
int main(int argc, char *argv[])
class OnlineTimer is used to test real-time decoding algorithms and evaluate how long the decoding of...
Definition: online-timing.h:88
This configuration class is to set up OnlineNnet2FeaturePipelineInfo, which in turn is the configurat...
For an extended explanation of the framework of which grammar-fsts are a part, please see Support for...
Definition: graph.dox:21
void PrintUsage(bool print_command_line=false)
Prints the usage documentation [provided in the constructor].
int32 g_num_threads
Definition: kaldi-thread.cc:25
This class stores the adaptation state from the online iVector extractor, which can help you to initi...
void OutputStats(OnlineTimingStats *stats)
This call, which should be made after decoding is done, writes the stats to the object that accumulat...
void SetBatchnormTestMode(bool test_mode, Nnet *nnet)
This function affects only components of type BatchNormComponent.
Definition: nnet-utils.cc:564
A templated class for writing objects to an archive or script file; see The Table concept...
Definition: kaldi-table.h:368
kaldi::int32 int32
BaseFloat SampFreq() const
Definition: wave-reader.h:126
const Matrix< BaseFloat > & Data() const
Definition: wave-reader.h:124
const Nnet & GetNnet() const
void Register(OptionsItf *opts)
void GetDiagnosticsAndPrintOutput(const std::string &utt, const fst::SymbolTable *word_syms, const CompactLattice &clat, int64 *tot_num_frames, double *tot_like)
bool GetLinearSymbolSequence(const Fst< Arc > &fst, std::vector< I > *isymbols_out, std::vector< I > *osymbols_out, typename Arc::Weight *tot_weight_out)
GetLinearSymbolSequence gets the symbol sequence from a linear FST.
This file contains a different version of the feature-extraction pipeline in online-feature-pipeline...
void Write(const std::string &key, const T &value) const
This class is responsible for storing configuration variables, objects and options for OnlineNnet2Fea...
void Read(std::istream &is, bool binary)
void Register(const std::string &name, bool *ptr, const std::string &doc)
void ReadKaldiObject(const std::string &filename, Matrix< float > *m)
Definition: kaldi-io.cc:832
This file contains some miscellaneous functions dealing with class Nnet.
Allows random access to a collection of objects in an archive or script file; see The Table concept...
Definition: kaldi-table.h:233
fst::LatticeWeightTpl< BaseFloat > LatticeWeight
Definition: kaldi-lattice.h:32
void CompactLatticeShortestPath(const CompactLattice &clat, CompactLattice *shortest_path)
A form of the shortest-path/best-path algorithm that&#39;s specially coded for CompactLattice.
void SetDropoutTestMode(bool test_mode, Nnet *nnet)
This function affects components of child-classes of RandomComponent.
Definition: nnet-utils.cc:573
std::vector< std::vector< double > > AcousticLatticeScale(double acwt)
std::istream & Stream()
Definition: kaldi-io.cc:826
float BaseFloat
Definition: kaldi-types.h:29
The class ParseOptions is for parsing command-line options; see Parsing command-line options for more...
Definition: parse-options.h:36
void Print(bool online=true)
Here, if "online == false" we take into account that the setup was used in not-really-online mode whe...
void ComputeCurrentTraceback(const LatticeFasterOnlineDecoderTpl< FST > &decoder)
const T & Value(const std::string &key)
void ScaleLattice(const std::vector< std::vector< ScaleFloat > > &scale, MutableFst< ArcTpl< Weight > > *fst)
Scales the pairs of weights in LatticeWeight or CompactLatticeWeight by viewing the pair (a...
void Read(std::istream &is, bool binary)
Struct OnlineCmvnState stores the state of CMVN adaptation between utterances (but not the state of t...
void ConvertLattice(const ExpandedFst< ArcTpl< Weight > > &ifst, MutableFst< ArcTpl< CompactLatticeWeightTpl< Weight, Int > > > *ofst, bool invert)
Convert lattice from a normal FST to a CompactLattice FST.
void GetLattice(bool end_of_utterance, CompactLattice *clat) const
Gets the lattice.
A templated class for reading objects sequentially from an archive or script file; see The Table conc...
Definition: kaldi-table.h:287
bool EndpointDetected(const OnlineEndpointConfig &config)
This function calls EndpointDetected from online-endpoint.h, with the required arguments.
You will instantiate this class when you want to decode a single utterance using the online-decoding ...
fst::VectorFst< LatticeArc > Lattice
Definition: kaldi-lattice.h:44
int Read(int argc, const char *const *argv)
Parses the command line options and fills the ParseOptions-registered variables.
#define KALDI_ERR
Definition: kaldi-error.h:147
GrammarFst is an FST that is &#39;stitched together&#39; from multiple FSTs, that can recursively incorporate...
Definition: grammar-fst.h:96
std::string GetArg(int param) const
Returns one of the positional parameters; 1-based indexing for argc/argv compatibility.
#define KALDI_WARN
Definition: kaldi-error.h:150
bool HasKey(const std::string &key)
fst::VectorFst< CompactLatticeArc > CompactLattice
Definition: kaldi-lattice.h:46
This class&#39;s purpose is to read in Wave files.
Definition: wave-reader.h:106
int NumArgs() const
Number of positional parameters (c.f. argc-1).
std::string global_cmvn_stats_rxfilename
Options for online cmvn, read from config file.
OnlineNnet2FeaturePipeline is a class that&#39;s responsible for putting together the various parts of th...
OnlineSilenceWeightingConfig silence_weighting_config
Config for weighting silence in iVector adaptation.
#define KALDI_VLOG(v)
Definition: kaldi-error.h:156
void AdvanceDecoding()
Advances the decoding as far as we can.
class OnlineTimingStats stores statistics from timing of online decoding, which will enable the Print...
Definition: online-timing.h:41
#define KALDI_LOG
Definition: kaldi-error.h:153
When you instantiate class DecodableNnetSimpleLooped, you should give it a const reference to this cl...
Represents a non-allocating general vector which can be defined as a sub-vector of higher-level vecto...
Definition: kaldi-vector.h:501
const LatticeFasterOnlineDecoderTpl< FST > & Decoder() const
void WaitUntil(double cur_utterance_length)
The call to WaitUntil(t) simulates the effect of sleeping until cur_utterance_length seconds after th...
void FinalizeDecoding()
Finalizes the decoding.
Config class for the CollapseModel function.
Definition: nnet-utils.h:240
void GetDeltaWeights(int32 num_frames_ready, int32 first_decoder_frame, std::vector< std::pair< int32, BaseFloat > > *delta_weights)