-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
351 lines (291 loc) · 14.4 KB
/
Copy pathindex.html
File metadata and controls
351 lines (291 loc) · 14.4 KB
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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description"
content="Pixel-GS: Density Control with Pixel-aware Gradient for 3D Gaussian Splatting">
<meta name="keywords" content="Pixel-GS: Density Control with Pixel-aware Gradient for 3D Gaussian Splatting">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Pixel-GS</title>
<!-- Bootstrap -->
<link rel="stylesheet" href="static/css/bootstrap-4.4.1.css">
<!-- Fonts -->
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet">
<!-- Bulma and Font Awesome -->
<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">
<link rel="icon" href="static/images/3.png">
<!-- jQuery and JS -->
<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>
<script src="static/js/app.js"></script>
<script src="static/js/video_comparison.js"></script>
<link rel="stylesheet" href="static/css/dics.original.css">
<script src="static/js/event_handler.js"></script>
<script src="static/js/dics.original.js"></script>
<script>
</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-1 publication-title" style="margin-bottom: 0"><strong>Pixel-GS</strong></h1>
<h2 class="title is-2 publication-title" style="margin-top: 0; margin-bottom: 0">Density Control with Pixel-aware Gradient for 3D Gaussian Splatting</h2>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="">Zheng Zhang</a><sup>1</sup></span>
<span class="author-block">
<a href="https://wbhu.github.io/">Wenbo Hu</a><sup>2+</sup>
</span>
<span class="author-block">
<a href="https://yxlao.github.io/">Yixing Lao</a><sup>1</sup></span>
<span class="author-block">
<a href="https://tonghe90.github.io/">Tong He</a><sup>3</sup>
</span>
<span class="author-block">
<a href="https://hszhao.github.io/">Hengshuang Zhao</a><sup>1+</sup>
</span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>1</sup>The University of Hong Kong</span>
<span class="author-block"><sup>2</sup>Tencent AI Lab</span>
<span class="author-block"><sup>3</sup>Shanghai AI Lab</span>
</div>
<div class="is-size-6 publication-authors">
<span class="author-block"><sup>+</sup>corresponding authors</span>
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- PDF Link. -->
<span class="link-block">
<a href="https://arxiv.org/pdf/2403.15530.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">
<a href="https://arxiv.org/abs/2403.15530"
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>
<!-- Code Link. -->
<span class="link-block">
<a href="https://github.com/zhengzhang01/Pixel-GS"
class="external-link button is-normal is-rounded is-dark disabled">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<div class="columns is-centered has-text-centered">
<div class="column is-full-width">
<video class="video" width="100%" id="xyalias6" loop playsinline autoplay muted src="resources/video/flowers/video.mp4" onplay="resizeAndPlay(this)" style="height: 0px;"></video>
<canvas height=0 class="videoMerge" id="xyalias6Merge"></canvas>
<h2 class="subtitle has-text-centered" style="margin-top: 15px">
<b>3DGS* (left)</b> tends to exhibit blurring and needle-like artifacts in areas with insufficient point cloud initialization, <br>
while <b>Pixel-GS (right)</b> by considering the conditions for point cloud growth from the pixel perspective effectively models these areas. <br>
<b>*</b> indicates retraining for better performance.
</h2>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3" style="margin-top: -30px">Abstract</h2>
<div class="content has-text-justified">
<p>
3D Gaussian Splatting (3DGS) has demonstrated impressive novel view synthesis results while advancing real-time rendering performance. However, it relies heavily on the quality of the initial point cloud, resulting in blurring and needle-like artifacts in areas with insufficient initializing points. This is mainly attributed to the point cloud growth condition in 3DGS that only considers the average gradient magnitude of points from observable views, thereby failing to grow for large Gaussians that are observable for many viewpoints while many of them are only covered in the boundaries. To this end, we propose a novel method, named Pixel-GS, to take into account the number of pixels covered by the Gaussian in each view during the computation of the growth condition. We regard the covered pixel numbers as the weights to dynamically average the gradients from different views, such that the growth of large Gaussians can be prompted. As a result, points within the areas with insufficient initializing points can be grown more effectively, leading to a more accurate and detailed reconstruction. Besides, we also propose a simple yet effective strategy to scale the gradient field according to the distance to the camera, to suppress the growth of floaters near the camera. Extensive experiments both qualitatively and quantitatively demonstrate that our method achieves state-of-the-art rendering quality while maintaining real-time rendering speed, on the challenging Mip-NeRF 360 and Tanks & Temples datasets.
</p>
</div>
</div>
</div>
</div>
</section>
<!DOCTYPE html>
<html>
<head>
<title>Your Page Title</title>
<script type="text/javascript" async
src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.5/MathJax.js?config=TeX-MML-AM_CHTML">
</script>
</head>
<body>
<section class="section">
<div class="container is-max-desktop">
<!-- Method. -->
<div class="columns is-centered has-text-centered">
<div class="column is-full-width">
<h2 class="title is-3" style="margin-top: -20px">Method</h2>
<img src="./resources/method.png" class="center">
<div class="content has-text-justified">
<p style="margin-top: 30px">
<b>Pipeline of Pixel-GS.</b> \( \mathrm{p}_i \) represents the number of pixels participating in the calculation for the Gaussian from this viewpoint, and \( \mathbf{g}_i \) represents the gradient of the Gaussian's NDC coordinates. We changed the condition for deciding whether a Gaussian should split or clone from the left to the right side.
</p>
</div>
</div>
</div>
<!--/ Animation. -->
</div>
</section>
</body>
</html>
<section class="section">
<div class="container is-max-desktop">
<!-- Results. -->
<div class="columns is-centered ">
<div class="column is-full-width">
<h2 class="title is-3 has-text-centered" style="margin-top: -30px">Results</h2>
<h3 class="title is-4">Comparison wtih 3DGS</h3>
<div class="content has-text-justified">
<p>
Here, we present the qualitative comparison results of 3DGS*, Pixel-GS (Ours), and Ground Truth.
</p>
</div>
<div class="container">
<ul class="nav nav-tabs nav-fill nav-justified" id="object-scale-recon">
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(0)">stump</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(1)">bicycle</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(2)">treehill</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(3)">flowers</a>
</li>
<li class="nav-item">
<a class="nav-link active" onclick="objectSceneEvent(4)">caterpillar</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(5)">barn</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(6)">playground</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(7)">train</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(8)">truck</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="objectSceneEvent(9)">m60</a>
</li>
</ul>
<div class="b-dics" style="width: 2600px; font-weight: 800;">
<img src="resources/image/caterpillar/3dgs.png" alt="3DGS*">
<img src="resources/image/caterpillar/ours.png" alt="Ours">
<img src="resources/image/caterpillar/gt.png" alt="GT">
</div>
</div>
<h3 class="title is-4" style="margin-top: 60px;">Drop a certain proportion of the initial point cloud.</h3>
<div class="content has-text-justified">
<p>
Randomly discard a certain proportion of the SFM point cloud points for scene initialization. The figure below shows the comparison of the reconstruction results for the bicycle scene from the Mip-NeRF 360 dataset between 3DGS, Pixel-GS (Ours) and Ground Truth.
</p>
</div>
<div class="container">
<ul class="nav nav-tabs nav-fill nav-justified" id="dropobject-scale-recon">
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(0)">00%</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(1)">10%</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(2)">30%</a>
</li>
<li class="nav-item">
<a class="nav-link active" onclick="dropobjectSceneEvent(3)">50%</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(4)">70%</a>
</li>
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(5)">90%</a>
</li>
</li>
<li class="nav-item">
<a class="nav-link" onclick="dropobjectSceneEvent(6)">99%</a>
</li>
</ul>
<div class="b-dics" style="width: 2600px; font-weight: 800;">
<img src="resources/drop/3dgs/50/00019.png" alt="3DGS*">
<img src="resources/drop/ours/50/00019.png" alt="Ours">
<img src="resources/drop/gt/50/00019.png" alt="GT">
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>@inproceedings{zhang2024pixelgs,
title = {Pixel-GS: Density Control with Pixel-aware Gradient for 3D Gaussian Splatting},
author = {Zhang, Zheng and Hu, Wenbo and Lao, Yixing and He, Tong and Zhao, Hengshuang},
booktitle = {ECCV},
year = {2024}
}</code></pre>
</div>
</section>
<footer class="footer">
<div class="container">
<div class="content has-text-centered">
</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>.
Website template credit to <a href="https://github.com/nerfies/nerfies.github.io">Nerfies</a>.
We also incorporated design elements from <a href="https://niujinshuchong.github.io/mip-splatting/">Mip-Splatting</a>.
The video comparison with sliding bar is from <a href="https://dorverbin.github.io/refnerf/">Ref-NeRF</a>.
The image comparison with sliding bar is from <a href="https://research.nvidia.com/labs/dir/neuralangelo/">Neuralangelo</a>.
</p>
</div>
</div>
<p></p>
</div>
</div>
</footer>
</body>
</html>