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
|
/*************************************************************************/
/* */
/* Centre for Speech Technology Research */
/* University of Edinburgh, UK */
/* Copyright (c) 1995,1996 */
/* All Rights Reserved. */
/* */
/* Permission is hereby granted, free of charge, to use and distribute */
/* this software and its documentation without restriction, including */
/* without limitation the rights to use, copy, modify, merge, publish, */
/* distribute, sublicense, and/or sell copies of this work, and to */
/* permit persons to whom this work is furnished to do so, subject to */
/* the following conditions: */
/* 1. The code must retain the above copyright notice, this list of */
/* conditions and the following disclaimer. */
/* 2. Any modifications must be clearly marked as such. */
/* 3. Original authors' names are not deleted. */
/* 4. The authors' names are not used to endorse or promote products */
/* derived from this software without specific prior written */
/* permission. */
/* */
/* THE UNIVERSITY OF EDINBURGH AND THE CONTRIBUTORS TO THIS WORK */
/* DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, INCLUDING */
/* ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO EVENT */
/* SHALL THE UNIVERSITY OF EDINBURGH NOR THE CONTRIBUTORS BE LIABLE */
/* FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES */
/* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN */
/* AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, */
/* ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF */
/* THIS SOFTWARE. */
/* */
/*************************************************************************/
/* Author : Simon King */
/* Date : July 1995 */
/*-----------------------------------------------------------------------*/
/* Compute delta coefficients for */
/* Tracks and Tracks */
/* */
/*=======================================================================*/
#include <cstdlib>
#include "EST_Track.h"
#include "EST_Wave.h"
# define MAX_DELTA_ORDER 2
/// max. number of points on which the delta co-eff is based
# define MAX_REGRESSION_LENGTH 4
static float compute_gradient(const EST_FVector &x, int num_points);
void delta(EST_Track &tr, EST_Track &d, int regression_length)
{
int reg_index, this_index;
// need at least two points to compute gradient
if ((regression_length < 2)||(regression_length > MAX_REGRESSION_LENGTH)){
cerr << "delta(EST_Track&, int) : ERROR : regression_length is "
<< regression_length << endl;
exit(0);
}
// temp stores the points passed to compute_gradient
EST_FVector temp(regression_length);
for (int j = 0; j < tr.num_channels(); j++ )
for (int i = 0; i < tr.num_frames(); i++)
{
// copy values needed to compute gradient into temp[]
for (reg_index=0; reg_index<regression_length; reg_index++)
{
// gradient is computed from points to left of current time
// rather than points centred around the current time
this_index = i - reg_index;
if (this_index >= 0)
temp[reg_index] = tr.a(this_index, j);
}
// gradient at frame 0 is defined as 0
if (i < 1)
d.a(i, j) = 0.0;
else if (i < regression_length - 1)
// enough data, but would prefer more
// number of points available is only i+1
d.a(i, j) = compute_gradient(temp, i + 1);
else
// plenty of data, use the last regression_length points
d.a(i, j) = compute_gradient(temp, regression_length);
}
}
void delta(EST_Wave &tr, EST_Wave &d, int regression_length)
{
int reg_index, this_index;
// need at least two points to compute gradient
if ((regression_length < 2)||(regression_length > MAX_REGRESSION_LENGTH)){
cerr << "delta(EST_Track&, int) : ERROR : regression_length is "
<< regression_length << endl;
exit(0);
}
// temp stores the points passed to compute_gradient
EST_FVector temp(regression_length);
for (int j = 0; j < tr.num_channels(); j++ )
for (int i = 0; i < tr.num_samples(); i++)
{
// copy values needed to compute gradient into temp[]
for (reg_index=0; reg_index<regression_length; reg_index++)
{
// gradient is computed from points to left of current time
// rather than points centred around the current time
this_index = i - reg_index;
if (this_index >= 0)
temp.a_no_check(reg_index) = (float)tr.a(this_index, j);
}
// gradient at frame 0 is defined as 0
if (i < 1)
d.a(i, j) = 0;
else if (i < regression_length - 1)
// enough data, but would prefer more
// number of points available is only i+1
d.a(i, j) = (short)compute_gradient(temp, i + 1);
else
// plenty of data, use the last regression_length points
d.a(i, j) = (short)compute_gradient(temp, regression_length);
}
}
static float compute_gradient(const EST_FVector &x, int num_points)
{
float gradient;
// NB x[0] is the point LATEST in time
// so x[1] is really x[t-1]
// time between points is assumed to be one unit
// These are solutions to least-squares fit of straight line
// to num_points points.
switch (num_points){
case 1:
gradient = 0.0;
break;
case 2:
gradient = x(0) - x(1);
break;
case 3:
gradient = (x(0) -x(2)) / 2.0;
break;
case 4:
gradient = ( (3.0*x(0)) + x(1) - x(2) - (3.0 * x(3)) ) / 10.0;
break;
default:
cerr << "compute_gradient(float*, int) : ERROR : num_points is"
<< num_points << endl;
exit(0);
break;
}
return gradient;
}
|
