get-full-lda-mat.cc
4.01 KB
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
// featbin/get-full-lda-mat.cc
// Copyright 2012-2013 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 "matrix/kaldi-matrix.h"
#include "transform/transform-common.h"
namespace kaldi {
void IncreaseTransformDimension(int32 new_dimension,
Matrix<BaseFloat> *mat) {
int32 d = mat->NumRows();
if (new_dimension < d)
KALDI_ERR << "--new-dimension argument invalid or not specified: "
<< new_dimension << " < " << d;
if (mat->NumCols() == d) { // linear transform d->d
mat->Resize(new_dimension, new_dimension, kCopyData);
for (int32 i = d; i < new_dimension; i++)
(*mat)(i, i) = 1.0; // set new dims to unit matrix.
} else if (mat->NumCols() == d+1) { // affine transform d->d.
Vector<BaseFloat> offset(mat->NumRows());
offset.CopyColFromMat(*mat, d);
mat->Resize(d, d, kCopyData); // remove offset from mat->
mat->Resize(new_dimension, new_dimension+1, kCopyData); // extend with zeros.
for (int32 i = d; i < new_dimension; i++)
(*mat)(i, i) = 1.0; // set new dims to unit matrix.
for (int32 i = 0; i < d; i++) // and set offset [last column]
(*mat)(d, i) = offset(i);
} else {
KALDI_ERR << "Input matrix has unexpected dimension " << d
<< " x " << mat->NumCols();
}
}
} // end namespace kaldi
int main(int argc, char *argv[]) {
try {
using namespace kaldi;
const char *usage =
"This is a special-purpose program to be used in \"predictive SGMMs\".\n"
"It takes in an LDA+MLLT matrix, and the original \"full\" LDA matrix\n"
"as output by the --write-full-matrix option of est-lda; and it writes\n"
"out a \"full\" LDA+MLLT matrix formed by the LDA+MLLT matrix plus the\n"
"remaining rows of the \"full\" LDA matrix; and also writes out its inverse\n"
"Usage: get-full-lda-mat [options] <lda-mllt-rxfilename> <full-lda-rxfilename> "
"<full-lda-mllt-wxfilename> [<inv-full-lda-mllt-wxfilename>]\n"
"E.g.: get-full-lda-mat final.mat full.mat full_lda_mllt.mat full_lda_mllt_inv.mat\n";
bool binary = true;
ParseOptions po(usage);
po.Register("binary", &binary, "Write in binary mode (only relevant if output is a wxfilename)");
po.Read(argc, argv);
if (po.NumArgs() < 3 || po.NumArgs() > 4) {
po.PrintUsage();
exit(1);
}
std::string lda_mllt_rxfilename = po.GetArg(1),
full_lda_rxfilename = po.GetArg(2),
full_lda_mllt_wxfilename = po.GetArg(3),
inv_full_lda_mllt_wxfilename = po.GetOptArg(4);
Matrix<BaseFloat> lda_mllt;
ReadKaldiObject(lda_mllt_rxfilename, &lda_mllt);
Matrix<BaseFloat> full_lda;
ReadKaldiObject(full_lda_rxfilename, &full_lda);
KALDI_ASSERT(full_lda.NumCols() == lda_mllt.NumCols());
KALDI_ASSERT(full_lda.NumRows() == full_lda.NumCols());
Matrix<BaseFloat> full_lda_mllt(full_lda);
full_lda_mllt.Range(0, lda_mllt.NumRows(),
0, lda_mllt.NumCols()).CopyFromMat(lda_mllt);
WriteKaldiObject(full_lda_mllt, full_lda_mllt_wxfilename, binary);
if (po.NumArgs() != 3) {
full_lda_mllt.Invert();
WriteKaldiObject(full_lda_mllt, inv_full_lda_mllt_wxfilename, binary);
}
return 0;
} catch(const std::exception &e) {
std::cerr << e.what();
return -1;
}
}