-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
428 lines (406 loc) · 15.9 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
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>
<meta charset="utf-8">
<meta name="description"
content="Object-Centric Domain Randomization for 3D Shape Reconstruction in the Wild">
<meta name="keywords" content="ObjectDR, Domain Randomization, DR, 3D Shape Reconstruction, 3D Object Reconstruction, in the wild, Out-of-Distribution Generalization, OOD, Domain Generalization, DG">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>ObjectDR</title>
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet">
<link rel="stylesheet" href="./static/css/bulma.min.css">
<link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="./static/css/bulma-slider.min.css">
<link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
<link rel="stylesheet"
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="./static/css/index.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="./static/js/fontawesome.all.min.js"></script>
<script src="./static/js/bulma-carousel.min.js"></script>
<script src="./static/js/bulma-slider.min.js"></script>
<script src="./static/js/index.js"></script>
</head>
<body>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-2 publication-title"><span class="is-size-3">[ ObjectDR ]</span> <br> Object-Centric Domain Randomization <br> for 3D Shape Reconstruction in the Wild</h1>
<div class="is-size-4 publication-authors" style="margin-top:-18px">
<span class="author-block" style="margin-right:12px">
<a href="https://jhcho99.github.io/">Junhyeong Cho</a><sup>1</sup></span>
<span class="author-block" style="margin-right:12px">
<a href="https://kim-youwang.github.io">Kim Youwang</a><sup>2</sup></span>
<span class="author-block" style="margin-right:12px">
<a href="https://scholar.google.co.kr/citations?user=mDxJj2AAAAAJ&hl=en">Hunmin Yang</a><sup>1,4</sup></span>
<span class="author-block">
<a href="https://ami.postech.ac.kr/members/tae-hyun-oh">Tae-Hyun Oh</a><sup>2,3,5</sup>
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block" style="margin-right:24px"><sup>1</sup>ADD</span>
<span class="author-block" style="margin-right:24px"><sup>2</sup>Department of EE, POSTECH</span>
<span class="author-block" style="margin-right:24px"><sup>3</sup>Graduate School of AI, POSTECH</span>
<span class="author-block"><sup>4</sup>KAIST</span>
<br>
<span class="author-block"><sup>5</sup>Institute for Convergence Research and Education in Advanced Technology, Yonsei University</span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- PDF Link. -->
<span class="link-block" style="margin-right:15px">
<a href="https://arxiv.org/pdf/2403.14539.pdf"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Paper</span>
</a>
</span>
<span class="link-block" style="margin-right:15px">
<a href="https://arxiv.org/abs/2403.14539"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>arXiv</span>
</a>
</span>
<span class="link-block" style="margin-right:15px">
<a href="#BibTeX"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-obp"></i>
</span>
<span>BibTex</span>
</a>
</span>
<span class="link-block">
<a href="mailto:ObjectDR.official@gmail.com"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-envelope"></i>
</span>
<span>Contact</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser" style="margin-top:-40px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Teaser.png" class="center"/>
</div>
<h2 class="subtitle has-text-centered" style="margin-top:-12px; margin-left:8px; margin-right:8px;">
The proposed data synthesis framework generates large-scale <b>⟨</b>3D shape, 2D image<b>⟩</b>-paired data via a random simulation of visual variations in object appearances and backgrounds.
<br><br>
<div class="gray-box-custom" style="margin-top:-12px">
To facilitate 3D shape reconstruction in the wild, we pre-train a model on our randomized data so that <span style="border-bottom: 2px solid; padding-bottom: 1.4px;"><b>it learns to capture a domain-invariant geometry prior</b></span> which is consistent across domains.
<br>
</div>
</h2>
</div>
</section>
<!-- Abstract. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Abstract</h2>
<div class="content has-text-justified">
<p>
One of the biggest challenges in single-view 3D shape reconstruction in the wild is the scarcity of <b>⟨</b>3D shape, 2D image<b>⟩</b>-paired data from real-world environments. Inspired by remarkable achievements via domain randomization, we propose <i>ObjectDR</i> which synthesizes such paired data via a random simulation of visual variations in object appearances and backgrounds.
</p>
<p>
Our data synthesis framework exploits a conditional generative model (e.g., ControlNet) to generate images conforming to spatial conditions such as 2.5D sketches, which are obtainable through a rendering process of 3D shapes from object collections (e.g., Objaverse-XL). To simulate diverse variations while preserving object silhouettes embedded in spatial conditions, we also introduce a disentangled framework which leverages an initial object guidance.
</p>
<p>
After synthesizing a wide range of data, we pre-train a model on them so that it learns to capture a domain-invariant geometry prior which is consistent across various domains. We validate its effectiveness by substantially improving 3D shape reconstruction models on a real-world benchmark. In a scale-up evaluation, our pre-training achieves 23.6% superior results compared with the pre-training on high-quality computer graphics renderings.
</p>
</div>
</div>
</div>
</section>
<br>
<br>
<section class="hero teaser" style="margin-top:-5px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table1.png" class="center-img"/>
</div>
</div>
</section>
<!-- Trade-off. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Diversity-Fidelity Trade-Off</h2>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:-5px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Trade-off.png" class="center-img"/>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<div class="content has-text-justified" style="margin-top:-22px">
<p style="margin-top:40px">
To effectively deal with the trade-off, we propose <b>ObjectDR<sub>dis</sub></b> which disentangles the randomization process of object appearances and backgrounds. The proposed framework randomizes object appearances and backgrounds in a separate manner, and then integrates them. Using this disentangled framework, we leverage an initial object guidance which significantly improves the fidelity at the expense of monotonous backgrounds. In parallel, we also enhance the diversity of backgrounds by synthesizing random authentic scenes without being constrained by objects.
</p>
</div>
</div>
</div>
</section>
<br>
<!-- ObjectDR. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">ObjectDR<sub>dis</sub></h2>
</div>
</div>
</section>
<!-- <br> -->
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Overview.png" class="center-img"/>
</div>
</div>
</section>
<br>
<!-- Synthesis Results. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Synthesis Results</h2>
</div>
</div>
<div class="is-centered has-text-centered content">
<p>
For the purpose of better visualizations, we draw bounding boxes with red rectangles.
</p>
</div>
</section>
<br>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Result1.png" class="center-img"/>
</div>
</div>
</section>
<section class="hero teaser" style="margin-top:-12px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Result2.png" class="center-img"/>
</div>
</div>
</section>
<section class="hero teaser" style="margin-top:-12px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Result3.png" class="center-img"/>
</div>
</div>
</section>
<!-- Experimental Results. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Experimental Results</h2>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:-5px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Figure4.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Figure5.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Figure6-7.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table2.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table3.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table4.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table5-6.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Table7.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Figure9.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/Figure10.png" class="center-img"/>
</div>
</div>
</section>
<br>
<!-- Discussion. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Discussion</h2>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/FigureE1.png" class="center-img"/>
</div>
</div>
</section>
<br>
<section class="hero teaser" style="margin-top:20px">
<div class="container is-max-desktop">
<div class="hero-body">
<img src="static/images/FigureE2.png" class="center-img"/>
</div>
</div>
</section>
<!-- Contact. -->
<hr>
<section class="section">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top:-5px">Contact</h2>
<div class="is-centered has-text-centered is-size-5">
<p>
ObjectDR (<a href="mailto:ObjectDR.official@gmail.com">ObjectDR.official@gmail.com</a>)
</p>
</div>
</div>
</div>
</section>
<!-- BibTex. -->
<hr>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">Citation</h2>
<pre><code>@article{cho2024objectdr,
title={Object-Centric Domain Randomization for 3D Shape Reconstruction in the Wild},
author={Junhyeong Cho and Kim Youwang and Hunmin Yang and Tae-Hyun Oh},
journal={arXiv preprint arXiv:2403.14539},
year={2024}
}</code></pre>
</div>
</section>
<br><br>
<footer class="footer">
<div class="container">
<div class="content has-text-centered">
<a class="icon-link"
href="https://arxiv.org/pdf/2403.14539.pdf">
<i class="fas fa-file-pdf"></i>
</a>
<a class="icon-link" href="https://jhcho99.github.io/" class="external-link" disabled>
<i class="fas fa-user"></i>
</a>
<a class="icon-link" href="mailto:ObjectDR.official@gmail.com" class="external-link" disabled>
<i class="fas fa-envelope"></i>
</a>
</div>
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This website is licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
<p>
This website template was adapted from <a href="https://nerfies.github.io/">Nerfies website</a>.
</p>
</div>
</div>
</div>
</div>
</footer>
</body>
</html>