-
Notifications
You must be signed in to change notification settings - Fork 143
/
Copy pathindex.htm
174 lines (149 loc) · 8.45 KB
/
index.htm
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
<!DOCTYPE html>
<html>
<!-- CSS and template comes from Bootstrap -->
<head>
<script src="cnn_fluids_files/jquery.js" type="text/javascript"></script>
<script src="cnn_fluids_files/blockui.js" type="text/javascript"></script>
<script src="cnn_fluids_files/highcharts.js"></script>
<script src="cnn_fluids_files/exporting.js"></script>
<script src="cnn_fluids_files/bootstrap.min.js"></script>
<script src="cnn_fluids_files/bootbox.min.js"></script>
<title>Accelerating Eulerian Fluid Simulation With Convolutional Networks</title>
<link rel="stylesheet" type="text/css" href="cnn_fluids_files/bootstrap.css">
</head>
<body>
<!-- <div class="container" style="width:1268px;"> -->
<div class="container" style="width:90%;" name="top_container">
<h1><table border="0" width="100%"><tbody><tr><td><a href="">Accelerating Eulerian Fluid Simulation With Convolutional Networks</a></td><td align="right"></td></tr></tbody></table></h1>
<div class="navbar">
<div class="navbar-inner">
<div class="container">
<ul class="nav">
<li class="active"><a class="menu_item" href="">Overview</a></li>
<li><a href="https://github.com/google/FluidNet" target="_blank">Code And Data</a></li>
<li><a class="menu_item" href="#references">References</a></li>
<li><a class="menu_item" href="#contact">Contact</a></li>
</ul>
</div>
</div>
</div>
<script type="text/javascript" charset="utf-8">
$(document).ready(function(){
$(".span12").hide();
$("#overview").show();
});
</script>
<div class="row" style="width:95%;">
<div class="span12" id="contact" style="width: 100%; display: none;">
<h3>Team</h3>
<table><tbody><tr>
<td style="border-right:solid 10px transparent;"><a href="https://https://jonathantompson.github.io/"><img src="cnn_fluids_files/jonathan.jpg" style="height: 140px"></a></td>
<td style="border-right:solid 10px transparent;"><a href="http://cims.nyu.edu/~schlacht/"><img src="cnn_fluids_files/kris.jpg" style="height: 140px"></a></td>
<td style="border-right:solid 10px transparent;"><a href="http://cims.nyu.edu/~pablo/"><img src="cnn_fluids_files/pablo.jpg" style="height: 140px"></a></td>
<td style="border-right:solid 10px transparent;"><a href="http://mrl.nyu.edu/~perlin/"><img src="cnn_fluids_files/ken.jpg" style="height: 140px"></a></td>
</tr>
<tr>
<td align="center" style="font-size:16px"><a href="https://jonathantompson.github.io/">Jonathan<br>Tompson</a></td>
<td align="center" style="font-size:16px"><a href="http://cims.nyu.edu/~schlacht/">Kristofer<br>Schlachter</a></td>
<td align="center" style="font-size:16px"><a href="http://cims.nyu.edu/~pablo/">Pablo<br>Sprechmann</a></td>
<td align="center" style="font-size:16px"><a href="http://mrl.nyu.edu/~perlin/">Ken<br>Perlin</a></td>
</tr></tbody></table>
<br>
<p><i class="icon-envelope icon-black"></i> For questions contact Jonathan Tompson: <i>[email protected]</i></p>
</div>
<div class="span12" id="references" style="width: 100%; display: none;">
<h3>References</h3>
<ul>
<li>
<p><a href="http://arxiv.org/abs/1607.03597"><strong>Accelerating Eulerian Fluid Simulation With Convolutional Networks</strong></a></p>
<p>Jonathan Tompson, Kristofer Schlachter, Pablo Sprechmann, Ken Perlin.</p>
<p>2016. Arxiv preprint.</p>
</li>
</ul>
</div>
<div class="span12" id="overview" style="width: 100%; display: block;">
<img src="cnn_fluids_files/sample.png" style="width: 100%" align="center">
<h3>Abstract</h3>
<p>
Real-time simulation of fluid and smoke is a long standing problem in computer graphics, where state-of-the-art approaches require large compute resources, making real-time applications often impractical. In this work, we propose a data-driven approach that leverages the approximation power of deep-learning methods with the precision of standard fluid solvers to obtain both fast and highly realistic simulations. The proposed method solves the incompressible Euler equations following the standard operator splitting method in which a large, often ill-condition linear system must be solved. We propose replacing this system by learning a Convolutional Network (ConvNet) from a training set of simulations using a semi-supervised learning method to minimize long-term velocity divergence.
</p>
<p>
ConvNets are amenable to efficient GPU implementations and, unlike exact iterative solvers, have fixed computational complexity and latency. The proposed hybrid approach restricts the learning task to a linear projection without modeling the well understood advection and body forces. We present real-time 2D and 3D simulations of fluids and smoke; the obtained results are realistic and show good generalization properties to unseen geometry.
</p>
<h3>Video</h3>
<table>
<tr>
<td align="center" style="font-size:16px"><a href="https://youtu.be/w71zxkniJfo">Demo Video</a></td>
</tr>
<tr>
<td><iframe width="560" height="315" src="https://www.youtube.com/embed/w71zxkniJfo" frameborder="0" allowfullscreen></iframe></rd>
</tr>
<!--
<tr>
<td align="center" style="font-size:16px"><a href="https://youtu.be/5rXd5-y3sEw">Smoke Column</a></td>
</tr>
<tr>
<td><iframe width="560" height="315" src="https://www.youtube.com/embed/5rXd5-y3sEw" frameborder="0" allowfullscreen></iframe></rd>
</tr>
<tr>
<td align="center" style="font-size:16px"><a href="https://youtu.be/_2sp60SZ-NM">Arch</a></td>
</tr>
<tr>
<td><iframe width="560" height="315" src="https://www.youtube.com/embed/_2sp60SZ-NM" frameborder="0" allowfullscreen></iframe></rd>
</tr>
<tr>
<td align="center" style="font-size:16px"><a href="https://youtu.be/MR4RHm4Qf50">Stanford Bunny</a></td>
</tr>
<tr>
<td><iframe width="560" height="315" src="https://www.youtube.com/embed/MR4RHm4Qf50" frameborder="0" allowfullscreen></iframe></rd>
</tr>
-->
</table>
<h3>Citing the dataset</h3>
<p>
<pre><code>@article{CNNFluid2016,
author = {{Tompson}, J. and {Schlachter}, K. and {Sprechmann}, P. and {Perlin}, K.},
title = "{Accelerating Eulerian Fluid Simulation With Convolutional Networks}",
journal = {ArXiv e-prints},
archivePrefix = "arXiv",
eprint = {1607.03597},
primaryClass = "cs.CV",
year = 2016,
month = jul,
}</code></pre></p>
<h3>Publication</h3>
<table>
<tr>
<td align="center"><a href="http://arxiv.org/abs/1607.03597"><img src="cnn_fluids_files/tog_paper_thumbnail.png" style="height: 120px;"></a></td>
<td> </rd>
<!--
<td align="center"><a href="http://cims.nyu.edu/~tompson/others/paper_presentation_no_videos.pptx"><img src="cnn_fluids_files/tog_ppt_thumbnail.jpg" style="height: 120px"></a></td>
-->
</tr>
<tr>
<td align="center" style="font-size:16px"><a href="http://arxiv.org/abs/1607.03597">arxiv preprint</a></td>
<td> </rd>
<!--
<td align="center" style="font-size:16px"><a href="http://cims.nyu.edu/~tompson/others/paper_presentation_no_videos.pptx">SIGGRAPH'14 ppt</a></td>
-->
</tr>
</table>
</div>
<div class="span12" id="download" style="width: 100%; display: block;">
<h3>Code and Data</h3>
<p>Coming soon! (contact <a href="mailto:[email protected]" target="_top">[email protected]</a> if you need immediate access to the codebase).</p>
</div>
</div>
<script>
<!--
$(".menu_item").click(function(){
$("ul.nav > li").removeClass("active");
$(".span12").hide();
theDiv = $(this).attr("href");
$(theDiv).show();
$(this).closest("li").toggleClass("active");
});
</script>
</div>
</body>
</html>