sgmm2-acc-stats2.cc

Go to the documentation of this file.

// sgmm2bin/sgmm2-acc-stats2.cc

// Copyright 2009-2012   Saarland University (Author:  Arnab Ghoshal),
//                       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.


#include "base/kaldi-common.h"
#include "util/common-utils.h"
#include "sgmm2/am-sgmm2.h"
#include "hmm/transition-model.h"
#include "sgmm2/estimate-am-sgmm2.h"
#include "hmm/posterior.h"

int main(int argc, char *argv[]) {
  using namespace kaldi;
  try {
    const char *usage =
        "Accumulate numerator and denominator stats for discriminative training\n"
        "of SGMMs (input is posteriors of mixed sign)\n"
        "Usage: sgmm2-acc-stats2 [options] <model-in> <feature-rspecifier> "
        "<posteriors-rspecifier> <num-stats-out> <den-stats-out>\n"
        "e.g.: sgmm2-acc-stats2 1.mdl 1.ali scp:train.scp ark:1.posts num.acc den.acc\n";

    ParseOptions po(usage);
    bool binary = true;
    std::string gselect_rspecifier, spkvecs_rspecifier, utt2spk_rspecifier;
    std::string update_flags_str = "vMNwucSt";
    BaseFloat rand_prune = 1.0e-05;

    po.Register("binary", &binary, "Write output in binary mode");
    po.Register("gselect", &gselect_rspecifier, "Precomputed Gaussian indices (rspecifier)");
    po.Register("spk-vecs", &spkvecs_rspecifier, "Speaker vectors (rspecifier)");
    po.Register("utt2spk", &utt2spk_rspecifier,
                "rspecifier for utterance to speaker map");
    po.Register("rand-prune", &rand_prune, "Pruning threshold for posteriors");
    po.Register("update-flags", &update_flags_str, "Which SGMM parameters to accumulate "
                "stats for: subset of vMNwcS.");

    po.Read(argc, argv);

    kaldi::SgmmUpdateFlagsType acc_flags = StringToSgmmUpdateFlags(update_flags_str);
    
    if (po.NumArgs() != 5) {
      po.PrintUsage();
      exit(1);
    }

    std::string model_filename = po.GetArg(1),
        feature_rspecifier = po.GetArg(2),
        posteriors_rspecifier = po.GetArg(3),
        num_accs_wxfilename = po.GetArg(4),
        den_accs_wxfilename = po.GetArg(5);
    

    using namespace kaldi;
    typedef kaldi::int32 int32;
    typedef kaldi::int64 int64;

    // Initialize the readers before the model, as the model can
    // be large, and we don't want to call fork() after reading it if
    // virtual memory may be low.
    SequentialBaseFloatMatrixReader feature_reader(feature_rspecifier);
    RandomAccessPosteriorReader posteriors_reader(posteriors_rspecifier);
    RandomAccessInt32VectorVectorReader gselect_reader(gselect_rspecifier);
    RandomAccessBaseFloatVectorReaderMapped spkvecs_reader(spkvecs_rspecifier,
                                                           utt2spk_rspecifier);
    RandomAccessTokenReader utt2spk_map(utt2spk_rspecifier);    
    
    AmSgmm2 am_sgmm;
    TransitionModel trans_model;
    {
      bool binary;
      Input ki(model_filename, &binary);
      trans_model.Read(ki.Stream(), binary);
      am_sgmm.Read(ki.Stream(), binary);
    }

    if (acc_flags & kSgmmSpeakerWeightProjections && !am_sgmm.HasSpeakerDependentWeights()) {
      acc_flags &= ~kSgmmSpeakerWeightProjections;
      KALDI_WARN << "Removing speaker weight projections (u) from flags "
          "as not present in model\n";
    }
    if (acc_flags & kSgmmSpeakerProjections && !am_sgmm.HasSpeakerSpace()) {
      acc_flags &= ~kSgmmSpeakerProjections;
      KALDI_WARN << "Removing speaker projections (N) from flags "
          "as not present in model\n";
    }
    
    Vector<double> num_transition_accs, den_transition_accs;
    if (acc_flags & kaldi::kSgmmTransitions) {
      trans_model.InitStats(&num_transition_accs);
      trans_model.InitStats(&den_transition_accs);
    }
    MleAmSgmm2Accs num_sgmm_accs(rand_prune), den_sgmm_accs(rand_prune);
    bool have_spk_vecs = (spkvecs_rspecifier != "");
    num_sgmm_accs.ResizeAccumulators(am_sgmm, acc_flags, have_spk_vecs);
    den_sgmm_accs.ResizeAccumulators(am_sgmm, acc_flags, have_spk_vecs);   

    double tot_like = 0.0, tot_weight = 0.0, tot_abs_weight = 0.0;
    int64 tot_frames = 0;

    kaldi::Sgmm2PerFrameDerivedVars per_frame_vars;

    int32 num_done = 0, num_err = 0;
    std::string cur_spk;
    Sgmm2PerSpkDerivedVars spk_vars;
    
    for (; !feature_reader.Done(); feature_reader.Next()) {
      std::string utt = feature_reader.Key();
      std::string spk = utt;
      if (!utt2spk_rspecifier.empty()) {
        if (!utt2spk_map.HasKey(utt)) {
          KALDI_WARN << "utt2spk map does not have value for " << utt
                     << ", ignoring this utterance.";
          continue;
        } else { spk = utt2spk_map.Value(utt); }
      }
      if (spk != cur_spk && cur_spk != "") {
        num_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
        den_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
      }
      if (spk != cur_spk || spk_vars.Empty()) {
        spk_vars.Clear();
        if (spkvecs_reader.IsOpen()) {
          if (spkvecs_reader.HasKey(utt)) {
            spk_vars.SetSpeakerVector(spkvecs_reader.Value(utt));
            am_sgmm.ComputePerSpkDerivedVars(&spk_vars);
          } else {
            KALDI_WARN << "Cannot find speaker vector for " << utt;
            num_err++;
            continue;
          }
        } // else spk_vars is "empty"
      }
      cur_spk = spk;
      
      const Matrix<BaseFloat> &features = feature_reader.Value();
      if (!posteriors_reader.HasKey(utt) ||
          posteriors_reader.Value(utt).size() != features.NumRows()) {
        KALDI_WARN << "No posterior info available for utterance "
                   << utt << " (or wrong size)";
        num_err++;
        continue;
      }
      
      const Posterior &posterior = posteriors_reader.Value(utt);
      if (!gselect_reader.HasKey(utt)
          && gselect_reader.Value(utt).size() != features.NumRows()) {
        KALDI_WARN << "No Gaussian-selection info available for utterance "
                   << utt << " (or wrong size)";
        num_err++;
      }
      const std::vector<std::vector<int32> > &gselect =
          gselect_reader.Value(utt);

      num_done++;
      BaseFloat tot_like_this_file = 0.0, tot_weight_this_file = 0.0,
          tot_abs_weight_this_file = 0.0;
        
      for (size_t i = 0; i < posterior.size(); i++) {
        if (posterior[i].empty())
          continue;
        am_sgmm.ComputePerFrameVars(features.Row(i), gselect[i], spk_vars,
                                    &per_frame_vars);
        
        for (size_t j = 0; j < posterior[i].size(); j++) {
          int32 tid = posterior[i][j].first,  // transition identifier.
              pdf_id = trans_model.TransitionIdToPdf(tid);
          BaseFloat weight = posterior[i][j].second,
              abs_weight = std::abs(weight);
            
          if (acc_flags & kaldi::kSgmmTransitions) {
            trans_model.Accumulate(abs_weight, tid,  weight > 0 ?
                                   &num_transition_accs : &den_transition_accs);
          }
          tot_like_this_file +=
              (weight > 0 ? num_sgmm_accs : den_sgmm_accs).Accumulate(
                  am_sgmm, per_frame_vars, pdf_id, abs_weight, &spk_vars)
              * weight;
          tot_weight_this_file += weight;
          tot_abs_weight_this_file += abs_weight;
        }
      }
      // Commit stats for the last speaker.
      num_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
      den_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
      
        
      tot_like += tot_like_this_file;
      tot_weight += tot_weight_this_file;
      tot_abs_weight += tot_abs_weight_this_file;
      tot_frames += posterior.size();
      if (num_done % 50 == 0)
        KALDI_LOG << "Processed " << num_done << " utterances.";
    }
    // Commit stats for last speaker.
    num_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
    den_sgmm_accs.CommitStatsForSpk(am_sgmm, spk_vars);
    
    KALDI_LOG << "Overall weighted acoustic likelihood per frame was "
              << (tot_like/tot_frames) << " over " << tot_frames << " frames; "
              << "average weight per frame is " << (tot_weight/tot_frames)
              << ", average abs(weight) per frame is "
              << (tot_abs_weight/tot_frames);
    
    KALDI_LOG << "Done " << num_done << " files, " << num_err
              << " with errors.";
    
    {
      Output ko(num_accs_wxfilename, binary);
      num_transition_accs.Write(ko.Stream(), binary);
      num_sgmm_accs.Write(ko.Stream(), binary);
    }
    {
      Output ko(den_accs_wxfilename, binary);
      den_transition_accs.Write(ko.Stream(), binary);
      den_sgmm_accs.Write(ko.Stream(), binary);
    }
    KALDI_LOG << "Written accs.";
    return (num_done != 0 ? 0 : 1);
  } catch(const std::exception &e) {
    std::cerr << e.what();
    return -1;
  }
}