-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
152 lines (135 loc) · 8.08 KB
/
index.html
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
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>Demo for "Attention-Guided Generative Adversarial Network for Whisper to Normal Speech Conversion"</title>
</head>
<body>
<h2 align="center"> Demo for "Attention-Guided Generative Adversarial Network for Whisper to Normal Speech Conversion"</h2>
<!-- <div><b>Authors:</b> Teng Gao, Jian Zhou, Huabin Wang, Liang Tao, and Hon Keung Kwan</div> -->
<div><b>Abstract:</b> Whispered speech is a special way of pronunciation without using vocal cord vibration. A whispered speech does not contain a fundamental frequency, and its energy is about 20dB lower than that of a normal speech. Converting a whispered speech into a normal speech can improve speech quality and intelligibility. In this paper, a novel attention-guided generative adversarial network model incorporating an autoencoder, a Siamese neural network, and an identity mapping loss function for whisper to normal speech conversion (AGAN-W2SC) is proposed. The proposed method avoids the challenge of estimating the fundamental frequency of the normal voiced speech converted from a whispered speech. Specifically, the proposed model is more amendable to practical applications because it does not need to align speech features for training. Experimental results demonstrate that the proposed AGAN-W2SC can obtain improved speech quality and intelligibility compared with dynamic-time-warping-based methods.</div>
<h4>Comparing of different models(Whisper, Normal, GMM, BLSTM, CycleGAN, Ours ): </h4>
<table border="1">
<tr>
<th>Whisper</th>
<th>Normal</th>
<!-- <th>lstm</th> -->
<th>GMM</th>
<th>BLSTM</th>
<th>CycleGAN</th>
<th>Ours</th>
<!-- <th>Causal-local-att-expanded</th> -->
<!-- <th>Causal-local-att-staecked</th> -->
</tr>
<tr>
<td align="center">A</td>
<td align="center">B</td>
<td align="center">C</td>
<td align="center">D</td>
<td align="center">E</td>
<td align="center">F</td>
<!-- <td align="center">F</td>
<td align="center">G</td>
<td align="center">H</td> -->
</tr>
</table>
<blockquote>
<table>
<tr>
<td colspan=2><span>1. test001.wav.</span></td>
</tr>
<tr>
<td>
<div style="float:left;width:20px; margin-top:16px;">A</div><audio controls>
<source src="Samples/Whisper/fw001.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">B</div><audio controls>
<source src="Samples/Normal/fn001.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">C</div><audio controls>
<source src="Samples/GMM/generated_fn001.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">D</div><audio controls>
<source src="Samples/BLSTM/generated_fn001.wav"></audio>
</td>
<!-- </tr>
<tr> -->
<td>
<div style="float:left;width:20px; margin-top:16px;">E</div><audio controls>
<source src="Samples/CycleGAN/generated_fn001.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">F</div><audio controls>
<source src="Samples/Ours/generated_fn001.wav"></audio>
</td>
<!-- <td><div style="float:left;width:20px; margin-top:16px;">F</div><audio controls><source src="samples/d-s/test001.wav"> </audio></td> <td><div style="float:left;width:20px; margin-top: 16px;">G</div><audio controls> <source src="samples/l-e/test001.wav"></audio></td><td><div style="float:left;width:20px; margin-top:16px;">H</div><audio controls><source src="samples/l-s/test001.wav"></audio></td></tr> -->
</tr>
<tr>
<td colspan=2><span>2. test002.wav.</span></td>
</tr>
<tr>
<td>
<div style="float:left;width:20px; margin-top:16px;">A</div><audio controls>
<source src="Samples/Whisper/fw002.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">B</div> <audio controls>
<source src="Samples/Normal/fn002.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">C</div> <audio controls>
<source src="Samples/GMM/generated_fn002.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">D</div> <audio controls>
<source src="Samples/BLSTM/generated_fn002.wav"></audio>
</td>
<!-- <tr> -->
<td>
<div style="float:left;width:20px; margin-top:16px;">E</div><audio controls>
<source src="Samples/CycleGAN/generated_fn002.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">F</div><audio controls>
<source src="Samples/Ours/generated_fn002.wav"></audio>
</td>
<!-- <td><div style="float:left;width:20px; margin-top:16px;">F</div><audio controls><source src="samples/d-s/test002.wav"></audio></td><td><div style="float:left;width:20px; margin-top:16px;">G</div><audio controls><source src="samples/l-e/test002.wav"></audio> </td><td><div style="float:left;width:20px; margin-top:16px;">H</div> <audio controls><source src="samples/l-s/test002.wav"></audio></td></tr> -->
</tr>
<tr>
<td colspan=2><span>3. test003.wav.</span></td>
</tr>
<tr>
<td>
<div style="float:left;width:20px; margin-top:16px;">A</div><audio controls>
<source src="Samples/Whisper/fw003.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">B</div><audio controls>
<source src="Samples/Normal/fn003.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">C</div><audio controls>
<source src="Samples/GMM/generated_fn003.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">D</div><audio controls>
<source src="Samples/BLSTM/generated_fn003.wav"></audio>
<!-- </tr>
<tr> -->
<td>
<div style="float:left;width:20px; margin-top:16px;">E</div><audio controls>
<source src="Samples/CycleGAN/generated_fn003.wav"></audio>
</td>
<td>
<div style="float:left;width:20px; margin-top:16px;">F</div><audio controls>
<source src="Samples/Ours/generated_fn003.wav"></audio>
</td>
<!-- <td><div style="float:left;width:20px; margin-top:16px;">F</div><audio controls><source src="samples/d-s/test003.wav"></audio></td><td><div style="float:left;width:20px; margin-top:16px;">G</div><audio controls><source src="samples/l-e/test003.wav"></audio></td><td><div style="float:left;width:20px; margin-top:16px;">H</div><audio controls><source src="samples/l-s/test003.wav"></audio></td></tr> -->
</tr>
</table>
</blockquote>
</body>
</html>