-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathindex.html
276 lines (230 loc) · 12.9 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
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta http-equiv="X-UA-Compatible" content="IE=edge">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>
DiffDub Project Page
</title>
<!-- bootstrap -->
<script src="https://code.jquery.com/jquery-3.4.1.slim.min.js" integrity="sha384-J6qa4849blE2+poT4WnyKhv5vZF5SrPo0iEjwBvKU7imGFAV0wwj1yYfoRSJoZ+n" crossorigin="anonymous"></script>
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/umd/popper.min.js" integrity="sha384-Q6E9RHvbIyZFJoft+2mJbHaEWldlvI9IOYy5n3zV9zzTtmI3UksdQRVvoxMfooAo" crossorigin="anonymous"></script>
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/css/bootstrap.min.css" integrity="sha384-Vkoo8x4CGsO3+Hhxv8T/Q5PaXtkKtu6ug5TOeNV6gBiFeWPGFN9MuhOf23Q9Ifjh" crossorigin="anonymous">
<script src="https://cdn.jsdelivr.net/npm/[email protected]/dist/js/bootstrap.min.js" integrity="sha384-wfSDF2E50Y2D1uUdj0O3uMBJnjuUD4Ih7YwaYd1iqfktj0Uod8GCExl3Og8ifwB6" crossorigin="anonymous"></script>
<!-- icon -->
<script src="https://kit.fontawesome.com/87dc3e863a.js" crossorigin="anonymous"></script>
<!-- font -->
<link href="https://fonts.googleapis.com/css?family=Open+Sans" rel="stylesheet" type="text/css">
<style>
body {
background: rgb(255, 255, 255) no-repeat fixed top left;
font-family:'Open Sans', sans-serif;
}
</style>
</head>
<body>
<!-- cover -->
<section>
<div class="jumbotron text-center mt-0" style="padding-bottom: 0px;">
<div class="container-fluid">
<div class="row">
<div class="col">
<!-- paper title -->
<h2 style="font-size:30px;">
DiffDub: Person-generic visual dubbing using inpainting renderer with diffusion auto-encoder
</h2>
<hr>
<!-- authors -->
<h6><a href = "mailto:[email protected]">Tao Liu</a><sup>1</sup>,
<a href = "mailto:[email protected]">Chenpeng Du</a><sup>1</sup>,
<a href = "mailto:[email protected]">Shuai Fan</a><sup>2</sup>,
<a href = "mailto:[email protected] ">Feilong Chen</a><sup>2</sup>,
<a href = "mailto:[email protected]">Kai Yu</a><sup>1</sup>,
<br>
<br>
<p> <sup>1</sup> MoE Key Lab of Artificial Intelligence, AI Institute, X-LANCE Lab, Shanghai Jiao Tong University <br>
<sup>2</sup> AISpeech Ltd, Suzhou China <br>
</p>
</h6>
<!-- links -->
<div class="row justify-content-center">
<!-- link to paper -->
<div style="margin: 5px;">
<p class="mb-5"><a class="btn btn-large btn-dark" href="https://arxiv.org/abs/2311.01811" role="button" target="_blank">
<i class="fa fa-file"></i>
Paper
</a>
</p>
</div>
<!-- link to the code -->
<div style="margin: 5px;">
<p class="mb-5"><a class="btn btn-large btn-dark" href=" https://github.com/liutaocode/DiffDub" role="button" target="_blank">
<i class="fa fa-file"></i>
Codes
</a>
</p>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- abstract -->
<section>
<div class="container">
<div class="row">
<div class="col-12 text-center">
<h3>Abstract</h3>
<hr style="margin-top:0px">
</div>
<p class="text-justify">
Generating high-quality and person-generic visual dubbing remains a challenge. Recent innovation has seen the advent of a two-stage paradigm, decoupling the rendering and lip synchronization process facilitated by intermediate representation as a conduit. Still, previous methodologies rely on rough landmarks or are confined to a single speaker, thus limiting their performance. In this paper, we propose DiffDub: Diffusion-based dubbing. We first craft the Diffusion auto-encoder by an inpainting renderer incorporating a mask to delineate editable zones and unaltered regions. This allows for seamless filling of the lower-face region while preserving the remaining parts. Throughout our experiments, we encountered several challenges. Primarily, the semantic encoder lacks robustness, constricting its ability to capture high-level features. Besides, the modeling ignored facial positioning, causing mouth or nose jitters across frames. To tackle these issues, we employ versatile strategies, including data augmentation and supplementary eye guidance. Moreover, we encapsulated a conformer-based reference encoder and motion generator fortified by a cross-attention mechanism. This enables our model to learn person-specific textures with varying references and reduces reliance on paired audio-visual data. Our rigorous experiments comprehensively highlight that our ground-breaking approach outpaces existing methods with considerable margins and delivers seamless, intelligible videos in person-generic and multilingual scenarios.
</p>
</div>
</div>
</div>
</section>
<br>
<br>
<!-- demo -->
<section>
<div class="container">
<div class="row">
<div class="col-12 text-center">
<h3>Method Comparsion on HDTF Reconstrution</h3>
<hr style="margin-top:0px">
</div>
<div >
<div>
<table border="0" style="text-align:center" width="100%">
<tr>
<td width="16.66%">Ground-truth</td>
<td width="16.66%">Wav2lip</td>
<td width="16.66%">PC-AVS</td>
<td width="16.66%">LP-LAP</td>
<td width="16.66%">DAE-Talker</td>
<td width="16.66%">Ours</td>
</tr>
</table>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/reconstruction/RD_Radio8_000_25fps.mp4" type="video/mp4"> </video> Video 1 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/reconstruction/RD_Radio22_000_25fps.mp4" type="video/mp4"> </video> Video 2 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/reconstruction/RD_Radio42_000_25fps.mp4" type="video/mp4"> </video> Video 3 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/reconstruction/RD_Radio51_000_25fps.mp4" type="video/mp4"> </video> Video 4 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/reconstruction/WRA_PeterRoskam0_000_25fps.mp4" type="video/mp4"> </video> Video 5 </div>
<a href="https://drive.google.com/drive/folders/1YGsG05ezgySJm9Iw7XeiVb_3FQRZLXFX?usp=sharing" target="_blank">More Videos </a>
</div>
</div>
</div>
</div>
<div class="container">
<div class="row">
<div class="col-12 text-center">
<h3>Method Comparsion on One-shot Dubbing</h3>
<hr style="margin-top:0px">
</div>
<div >
<div>
<table border="0" style="text-align:center" width="100%">
<tr>
<td width="16.66%">Portrait</td>
<td width="16.66%">Wav2lip</td>
<td width="16.66%">PC-AVS</td>
<td width="16.66%">LP-LAP</td>
<td width="16.66%">DAE-Talker</td>
<td width="16.66%">Ours</td>
</tr>
</table>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/one_shot/RD_Radio5_000_25fps.mp4" type="video/mp4"> </video> Video 1 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/one_shot/RD_Radio11_000_25fps.mp4" type="video/mp4"> </video> Video 2 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/one_shot/RD_Radio17_000_25fps.mp4" type="video/mp4"> </video> Video 3 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/one_shot/RD_Radio25_000_25fps.mp4" type="video/mp4"> </video> Video 4 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/one_shot/RD_Radio29_000_25fps.mp4" type="video/mp4"> </video> Video 5 </div>
<a href="https://drive.google.com/drive/folders/1YGsG05ezgySJm9Iw7XeiVb_3FQRZLXFX?usp=sharing" target="_blank">More Videos </a>
</div>
</div>
</div>
</div>
<div class="container">
<div class="row">
<div class="col-12 text-center">
<h3>Method Comparsion on Few-shot Dubbing</h3>
<hr style="margin-top:0px">
</div>
<div >
<div>
<table border="0" style="text-align:center" width="100%">
<tr>
<td width="20%">Wav2lip</td>
<td width="20%">PC-AVS</td>
<td width="20%">LP-LAP</td>
<td width="20%">DAE-Talker</td>
<td width="20%">Ours</td>
</tr>
</table>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/few_shot/test3_d6.mp4" type="video/mp4"> </video> Video 1 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/few_shot/test5_driving_audio_2.mp4" type="video/mp4"> </video> Video 2 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/few_shot/test6_driving_audio_5.mp4" type="video/mp4"> </video> Video 3 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/few_shot/test12_d8.mp4" type="video/mp4"> </video> Video 4 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/few_shot/test19_driving_audio_6.mp4" type="video/mp4"> </video> Video 5 </div>
<a href="https://drive.google.com/drive/folders/1YGsG05ezgySJm9Iw7XeiVb_3FQRZLXFX?usp=sharing" target="_blank">More Videos </a>
</div>
</div>
</div>
</div>
<div class="container">
<div class="row">
<div class="col-12 text-center">
<h3>Ablation Study</h3>
<hr style="margin-top:0px">
</div>
<div >
<div>
<table border="0" style="text-align:center" width="100%">
<tr>
<td width="16.66%">Ground-truth</td>
<td width="16.66%">w/o eye gudiance</td>
<td width="16.66%">w/o image aug.</td>
<td width="16.66%">w/o w.s.</td>
<td width="16.66%">10% paired data</td>
<td width="16.66%">Ours</td>
</tr>
</table>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/ablation_study/RD_Radio4_000_25fps.mp4" type="video/mp4"> </video> Video 1 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/ablation_study/RD_Radio8_000_25fps.mp4" type="video/mp4"> </video> Video 2 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/ablation_study/RD_Radio13_000_25fps.mp4" type="video/mp4"> </video> Video 3 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/ablation_study/RD_Radio18_000_25fps.mp4" type="video/mp4"> </video> Video 4 </div>
<div style="text-align:center"> <video width="100%" controls> <source src="resources/ablation_study/RD_Radio21_000_25fps.mp4" type="video/mp4"> </video> Video 5 </div>
<a href="https://drive.google.com/drive/folders/1YGsG05ezgySJm9Iw7XeiVb_3FQRZLXFX?usp=sharing" target="_blank">More Videos </a>
</div>
<br>
<b>Statement</b>: The video sources on this page are from <a href="https://github.com/universome/HDTF">HDTF</a> and <a href="https://github.com/MRzzm/DINet">DINet</a>, and are solely intended for academic demonstration purposes without any other offensive intent.
</div>
</div>
</div>
</div>
</div>
</section>
<br>
<br>
<!-- citing -->
<div class="container">
<div class="row ">
<div class="col-12">
<h3>Citation</h3>
<hr style="margin-top:0px">
<pre style="background-color: #e9eeef;padding: 1.25em 1.5em">
<code>@inproceedings{liu2024diffdub,
title={DiffDub: Person-Generic Visual Dubbing Using Inpainting Renderer with Diffusion Auto-Encoder},
author={Liu, Tao and Du, Chenpeng and Fan, Shuai and Chen, Feilong and Yu, Kai},
booktitle={ICASSP 2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={3630--3634},
year={2024},
organization={IEEE}
}</code></pre>
</div>
</div>
</div>
</body>
</html>