-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
428 lines (375 loc) · 18.7 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
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
<!DOCTYPE html>
<html>
<head lang="en">
<meta charset="UTF-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<title>Export3D</title>
<meta name="description" content="">
<meta name="viewport" content="width=device-width, initial-scale=1">
<!-- <base href="/"> -->
<!--FACEBOOK-->
<meta property="og:image" content="images/overview.png">
<meta property="og:image:type" content="image/png">
<meta property="og:image:width" content="682">
<meta property="og:image:height" content="682">
<meta property="og:type" content="website" />
<meta property="og:url" content="https://export3d.github.io"/>
<meta property="og:title" content="ExPort3D" />
<meta property="og:description" content="Project page for Export3D: Learning to Generate Conditional Tri-plane for 3D-aware Expression Controllable Portrait Animation" />
<!--TWITTER-->
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="Export3D" />
<meta name="twitter:description" content="Project page for Export3D: Learning to Generate Conditional Tri-plane for 3D-aware Expression Controllable Portrait Animation" />
<meta name="twitter:image" content="images/overview.png" />
<!-- <link rel="apple-touch-icon" href="apple-touch-icon.png"> -->
<!-- <link rel="icon" type="image/png" href="img/seal_icon.png"> -->
<!-- Place favicon.ico in the root directory -->
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/css/bootstrap.min.css">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/font-awesome/4.4.0/css/font-awesome.min.css">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/codemirror/5.8.0/codemirror.min.css">
<link rel="stylesheet" href="css/app.css">
<link rel="stylesheet" href="css/bootstrap.min.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/js/bootstrap.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/codemirror/5.8.0/codemirror.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/clipboard.js/1.5.3/clipboard.min.js"></script>
<script src="js/app.js"></script>
</head>
<body>
<div class="container" id="main">
<div class="row">
<h2 class="col-md-12 text-center">
<b>Export3D</b>:
Learning to Generate Conditional Tri-plane for</br> 3D-aware Expression Controllable Portrait Animation</br>
<b> <small>ECCV 2024</small> </b>
</h2>
</div>
<div class="row">
<div class="col-md-12 text-center">
<ul class="list-inline">
<li>
<a href="https://taekyungki.github.io">
Taekyung Ki
</a>
</br> DeepBrain AI Inc.
</li>
<li>
<a href="https://kevinmin95.github.io">
Dongchan Min
</a>
</br> Graduate School of AI, KAIST
</li>
<li>
<a href="https://www.deepbrain.io">
Gyeongsu Chae
</a>
</br> DeepBrain AI Inc.
</li>
</li><br>
</ul>
<ul class="'list-inline">
</ul>
</div>
</div>
<div class="row">
<div class="col-md-4 col-md-offset-4 text-center">
<ul class="nav nav-pills nav-justified">
<li>
<a href="https://arxiv.org/abs/2404.00636">
<image src="images/paper.png" height="60px">
<h5><strong>Paper</strong></h5>
</a>
</li>
<li>
<image src="images/youtube_icon.png" height="60px">
<h5><strong>Video (N/A) </strong></h5>
</li>
</ul>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
Abstract
</h3>
<image src="images/overview.png" class="img-responsive" alt="overview"><br>
<p class="text-justify">
In this paper, we present Export3D, a one-shot 3D-aware portrait animation method that is able to control the facial expression and camera view of a given portrait image.
To achieve this, we introduce a tri-plane generator that directly generates a tri-plane of 3D prior by transferring the expression parameter of 3DMM into the source image.
The tri-plane is then decoded into the image of different view through a differentiable volume rendering. Existing portrait animation methods heavily rely on image warping to transfer the expression in the motion space, challenging on disentanglement of appearance and expression.
In contrast, we propose a contrastive pre-training framework for appearance-free expression parameter,
eliminating undesirable appearance swap when transferring a cross-identity expression. Extensive experiments show that our pre-training framework can learn the appearance-free expression representation hidden in 3DMM,
and our model can generate <strong>3D</strong>-aware <strong>ex</strong>pression controllable <strong>port</strong>rait image without appearance swap in the cross-identity manner.
</p>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
One-shot Novel-view Animation Results
</h3>
<p class="text-justify">
Our method can generate 3D-aware videos using a single source image. Here, we fix the expression and
animate the source (the first row) by controlling the camera parameters (the second row).
</p>
<video id="v0" width="100%" autoplay muted loop controls>
<source src="videos/free-view_vox1.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 1.</b> One-shot free view animation results on Voxceleb1 test data.
<br>
</p>
<video id="v0" width="100%" autoplay muted loop controls>
<source src="videos/free-view_vfhq.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 2.</b> One-shot free view animation results on VFHQ test data.
<br><br>
</p>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
One-shot Reconstruction Results
</h3>
<p class="text-justify">
Our method can faithfully reconstruct facial expressions (e.g., eye-blink, mouth motion) and head pose using a single reference image.
Here, the source images are the same as above. Note that for intuitive comparison, we incorporate the corresponding audio into the output videos.
<br>First column: Groud Truth. Second column: Generated Result. Third column: Estimated Depth.
</p>
<video id="v0" width="49%" controls muted>
<source src="videos/id10282_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id00001_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id10283_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id00002_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id10284_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id00005_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id10292_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<video id="v0" width="49%" controls muted>
<source src="videos/id00007_recon-w-gt-depth.mp4" type="video/mp4" />
</video>
<p class="text-justify ">
<b>Video 3-6.</b> One-shot Reconstruction on Voxceleb1 test.
<b>Video 7-10.</b> One-shot Reconstruction on VFHQ test.
<br><br>
</p>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
One-shot Cross Identity Transfer Results
</h3>
<p class="text-justify">
We provide cross expression+camera results with different identities in VFHQ.
Our method can preserve the visual identity only with a single reference image (the first row) and the expression of the driving video (the first column).
</p>
<video id="v0" width="100%" controls muted>
<source src="videos/1-cce-w-gt.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 11.</b> One-shot Cross Expression+Camera on VFHQ test data (1).
<br>
</p>
<video id="v0" width="100%" controls muted>
<source src="videos/2-cce-w-gt.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 12.</b> One-shot Cross Expression+Camera on VFHQ test data (2).
<br>
</p>
<br><br>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
Comparison with other Methods
</h3>
<h4>
Same-identity Transfer Results
</h4>
<video id="v0" width="100%" controls>
<source src="videos/1a_compare_same.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 13a.</b> Same-identity Comparision in VFHQ.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/1b_compare_same.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 13b.</b> Same-identity Comparision in TalkingHead-1KH.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/1c_compare_same.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 13c.</b> Same-identity Comparision in VFHQ.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/1d_compare_same.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 13d.</b> Same-identity Comparision in TalkingHead-1KH.
<br>
</p>
<br>
<h4>
Cross-identity Transfer Results
</h4>
<video id="v0" width="100%" controls>
<source src="videos/2a_compare_cross.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 14a.</b> Cross-identity Comparision in VFHQ.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/2b_compare_cross.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 14b.</b> Cross-identity Comparision in TalkingHead-1KH.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/2c_compare_cross.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 14c.</b> Cross-identity Comparision in VFHQ.
<br>
</p>
<video id="v0" width="100%" controls>
<source src="videos/2d_compare_cross.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 14d.</b> Cross-identity Comparision in TalkingHead-1KH.
<br>
</p>
<br><br>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
About Expression Controls
</h3>
<center>
<image src="images/3dmm-param_vs_clebs-param.png" class="img-responsive" alt="tsne" width="80%"><br>
<p class="text-justify">
To explicitly control the facial expression of a given source image, we leverage 3DMM expression parameters.
However, we observe that raw 3DMM expression parameters still contain the appearance information even in the low-dimensional space (t-SNE), resulting in undesirable appearance swaps, such as eye shape, head size, etc., when transferring cross-identity expression.
To alleviate this issue, we propose a <strong>contrastive pre-training framework to extract the expression information hidden in the expression parameters</strong>.
Further details about the framework are in our paper.
</p>
</center>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/3c_ablation.mp4" type="video/mp4" />
</video>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/3d_ablation.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 15.</b> Ablation Studies on Expression Encoding.
<br>
</p>
<br>
<h4>
Expression Control along the Orthogonal Directions
</h4>
<center>
<p class="text-justify">
At the same time, we enforce the orthogonal structure on them, representing different expressions along different basis vectors.
</p>
</center>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/4a_linear-v1-v5.mp4" type="video/mp4" />
</video>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/4b_linear-v6-v10.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 16.</b> Visualization of Expression Control along Orthogonal Directions.
<br>
</p>
<br><br>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
Novel-view Synthesis Results with Expression Transfer
</h3>
<center>
<p class="text-justify">
Our method can transfer the driving expression to the source image without any multi-view inconsistency in terms of lighting and texture.
</p>
</center>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/5a_novel_view.mp4" type="video/mp4" />
</video>
<video id="v0" width="100%" autoplay controls muted loop>
<source src="videos/5b_novel_view.mp4" type="video/mp4" />
</video>
<p class="text-justify">
<b>Video 17.</b> Results of Novel-view Synthesis with Expression Transfer.
<br>
</p>
<br><br>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
Citation
</h3>
<div class="form-group col-md-10 col-md-offset-1">
<textarea id="bibtex" class="form-control" readonly>
@article{ki2024learning,
title={Learning to Generate Conditional Tri-plane for 3D-aware Expression Controllable Portrait Animation},
author={Ki, Taekyung and Min, Dongchan and Chae, Gyeongsu},
journal={arXiv preprint arXiv:2404.00636},
year={2024}
}</textarea>
</div>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
Related Links
</h3>
<p class="text-justify">
<a href="https://github.com/NVlabs/eg3d">EG3D</a> provides an excellent code base for our work.
</p>
<p class="text-justify">
The website template was borrowed from <a href="http://mgharbi.com/">Michaël Gharbi</a>, and <a href="https://jonbarron.info/mipnerf/"> Mip-NeRF</a>.
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h3>
</h3>
</p>
</div>
</div>
</div>
</body>
</html>