// nnet2bin/nnet-perturb-egs.cc

// Copyright 2012-2014 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.

#include "base/kaldi-common.h"

#include "util/common-utils.h"

#include "hmm/transition-model.h"

#include "nnet2/nnet-example-functions.h"

namespace kaldi {

namespace nnet2 {

void PerturbTrainingExample(const TpMatrix<BaseFloat> &cholesky,

BaseFloat noise_factor,

NnetExample *eg) {

Matrix<BaseFloat> input_frames(eg->input_frames);

int32 dim = input_frames.NumRows() * input_frames.NumCols();

if (dim != cholesky.NumRows()) {

KALDI_ERR << "Dimension mismatch: egs have total dim " << dim

<< " vs. cholesky factor " << cholesky.NumRows();

}

Vector<BaseFloat> vec(dim, kUndefined);

vec.CopyRowsFromMat(input_frames);

Vector<BaseFloat> noise(dim);

noise.SetRandn(); // Gaussian noise with unit variance and zero mean

vec.AddTpVec(noise_factor, cholesky, kNoTrans, noise, 1.0);

input_frames.CopyRowsFromVec(vec);

eg->input_frames.CopyFromMat(input_frames);

}

}

}

int main(int argc, char *argv[]) {

try {

using namespace kaldi;

using namespace kaldi::nnet2;

typedef kaldi::int32 int32;

typedef kaldi::int64 int64;

const char *usage =

"Copy examples, perturbing them by adding a specified amount (--noise-factor)\n"

"times the within-class covariance of the examples. the Cholesky factor of\n"

"the examples (obtained from the --write-cholesky option of\n"

"nnet-get-feature-transform) must be supplied.\n"

"\n"

"Usage: nnet-perturb-egs [options] <cholesky> <egs-rspecifier> <egs-wspecifier>\n"

"\n"

"nnet-perturb-egs --noise-factor=0.2 exp/nnet5/cholesky.tpmat ark:- ark:-\n";

BaseFloat noise_factor = 0.1;

int32 srand_seed = 0;

ParseOptions po(usage);

po.Register("noise-factor", &noise_factor, "Factor to multiply noise generated "

"from within-class variance by before adding to egs");

po.Register("srand", &srand_seed, "Seed for random number generator ");

po.Read(argc, argv);

srand(srand_seed);

if (po.NumArgs() != 3) {

po.PrintUsage();

exit(1);

}

std::string cholesky_rxfilename = po.GetArg(1),

examples_rspecifier = po.GetArg(2),

examples_wspecifier = po.GetArg(3);

TpMatrix<BaseFloat> cholesky;

ReadKaldiObject(cholesky_rxfilename, &cholesky);

SequentialNnetExampleReader example_reader(examples_rspecifier);

NnetExampleWriter example_writer(examples_wspecifier);

int64 num_done = 0;

for (; !example_reader.Done(); example_reader.Next(), num_done++) {

std::string key = example_reader.Key();

NnetExample eg = example_reader.Value();

PerturbTrainingExample(cholesky, noise_factor, &eg);

example_writer.Write(key, eg);

}

KALDI_LOG << "Perturbed " << num_done << " neural-network training examples "

<< "with noise factor " << noise_factor;

return (num_done == 0 ? 1 : 0);

} catch(const std::exception &e) {

std::cerr << e.what() << '\n';

return -1;

}

}