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
|
/*
* Copyright (c) 2012 The WebRTC project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <stdlib.h>
#include "common_audio/vad/vad_unittest.h"
#include "test/gtest.h"
extern "C" {
#include "common_audio/vad/vad_core.h"
#include "common_audio/vad/vad_filterbank.h"
}
namespace webrtc {
namespace test {
const int kNumValidFrameLengths = 3;
TEST_F(VadTest, vad_filterbank) {
VadInstT* self = reinterpret_cast<VadInstT*>(malloc(sizeof(VadInstT)));
static const int16_t kReference[kNumValidFrameLengths] = {48, 11, 11};
static const int16_t kFeatures[kNumValidFrameLengths * kNumChannels] = {
1213, 759, 587, 462, 434, 272, 1479, 1385, 1291,
1200, 1103, 1099, 1732, 1692, 1681, 1629, 1436, 1436};
static const int16_t kOffsetVector[kNumChannels] = {368, 368, 272,
176, 176, 176};
int16_t features[kNumChannels];
// Construct a speech signal that will trigger the VAD in all modes. It is
// known that (i * i) will wrap around, but that doesn't matter in this case.
int16_t speech[kMaxFrameLength];
for (size_t i = 0; i < kMaxFrameLength; ++i) {
speech[i] = static_cast<int16_t>(i * i);
}
int frame_length_index = 0;
ASSERT_EQ(0, WebRtcVad_InitCore(self));
for (size_t j = 0; j < kFrameLengthsSize; ++j) {
if (ValidRatesAndFrameLengths(8000, kFrameLengths[j])) {
EXPECT_EQ(kReference[frame_length_index],
WebRtcVad_CalculateFeatures(self, speech, kFrameLengths[j],
features));
for (int k = 0; k < kNumChannels; ++k) {
EXPECT_EQ(kFeatures[k + frame_length_index * kNumChannels],
features[k]);
}
frame_length_index++;
}
}
EXPECT_EQ(kNumValidFrameLengths, frame_length_index);
// Verify that all zeros in gives kOffsetVector out.
memset(speech, 0, sizeof(speech));
ASSERT_EQ(0, WebRtcVad_InitCore(self));
for (size_t j = 0; j < kFrameLengthsSize; ++j) {
if (ValidRatesAndFrameLengths(8000, kFrameLengths[j])) {
EXPECT_EQ(0, WebRtcVad_CalculateFeatures(self, speech, kFrameLengths[j],
features));
for (int k = 0; k < kNumChannels; ++k) {
EXPECT_EQ(kOffsetVector[k], features[k]);
}
}
}
// Verify that all ones in gives kOffsetVector out. Any other constant input
// will have a small impact in the sub bands.
for (size_t i = 0; i < kMaxFrameLength; ++i) {
speech[i] = 1;
}
for (size_t j = 0; j < kFrameLengthsSize; ++j) {
if (ValidRatesAndFrameLengths(8000, kFrameLengths[j])) {
ASSERT_EQ(0, WebRtcVad_InitCore(self));
EXPECT_EQ(0, WebRtcVad_CalculateFeatures(self, speech, kFrameLengths[j],
features));
for (int k = 0; k < kNumChannels; ++k) {
EXPECT_EQ(kOffsetVector[k], features[k]);
}
}
}
free(self);
}
} // namespace test
} // namespace webrtc
|
