Skip to content

Commit

Permalink
Update index.html
Browse files Browse the repository at this point in the history
  • Loading branch information
@MarcTLaw
MarcTLaw committed Oct 23, 2023
1 parent 3648b86 commit edd16c5
Showing 1 changed file with 5 additions and 390 deletions.
395 changes: 5 additions & 390 deletions docs/index.html
View file
Original file line number Diff line number Diff line change
@@ -1,390 +1,5 @@
<head>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-137506474-1"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());

gtag('config', 'UA-137506474-1');
</script>

<script src="http://www.google.com/jsapi" type="text/javascript"></script>
<script type="text/javascript">google.load("jquery", "1.3.2");</script>
</head>

<style type="text/css">
body {
font-family: "HelveticaNeue-Light", "Helvetica Neue Light", "Helvetica Neue", Helvetica, Arial, "Lucida Grande", sans-serif;
font-weight:300;
font-size:18px;
margin-left: auto;
margin-right: auto;
width: 1100px;
}

h1 {
font-weight:300;
margin: 0.4em;
}

p {
margin: 0.2em;
}

.disclaimerbox {
background-color: #eee;
border: 1px solid #eeeeee;
border-radius: 10px ;
-moz-border-radius: 10px ;
-webkit-border-radius: 10px ;
padding: 20px;
}

video.header-vid {
height: 140px;
border: 1px solid black;
border-radius: 10px ;
-moz-border-radius: 10px ;
-webkit-border-radius: 10px ;
}

img.header-img {
height: 140px;
border: 1px solid black;
border-radius: 10px ;
-moz-border-radius: 10px ;
-webkit-border-radius: 10px ;
}

img.rounded {
border: 1px solid #eeeeee;
border-radius: 10px ;
-moz-border-radius: 10px ;
-webkit-border-radius: 10px ;
}

a:link,a:visited
{
color: #1367a7;
text-decoration: none;
}
a:hover {
color: #208799;
}

td.dl-link {
height: 160px;
text-align: center;
font-size: 22px;
}

.layered-paper-big { /* modified from: http://css-tricks.com/snippets/css/layered-paper/ */
box-shadow:
0px 0px 1px 1px rgba(0,0,0,0.35), /* The top layer shadow */
5px 5px 0 0px #fff, /* The second layer */
5px 5px 1px 1px rgba(0,0,0,0.35), /* The second layer shadow */
10px 10px 0 0px #fff, /* The third layer */
10px 10px 1px 1px rgba(0,0,0,0.35), /* The third layer shadow */
15px 15px 0 0px #fff, /* The fourth layer */
15px 15px 1px 1px rgba(0,0,0,0.35), /* The fourth layer shadow */
20px 20px 0 0px #fff, /* The fifth layer */
20px 20px 1px 1px rgba(0,0,0,0.35), /* The fifth layer shadow */
25px 25px 0 0px #fff, /* The fifth layer */
25px 25px 1px 1px rgba(0,0,0,0.35); /* The fifth layer shadow */
margin-left: 10px;
margin-right: 45px;
}


.layered-paper { /* modified from: http://css-tricks.com/snippets/css/layered-paper/ */
box-shadow:
0px 0px 1px 1px rgba(0,0,0,0.35), /* The top layer shadow */
5px 5px 0 0px #fff, /* The second layer */
5px 5px 1px 1px rgba(0,0,0,0.35), /* The second layer shadow */
10px 10px 0 0px #fff, /* The third layer */
10px 10px 1px 1px rgba(0,0,0,0.35); /* The third layer shadow */
margin-top: 5px;
margin-left: 10px;
margin-right: 30px;
margin-bottom: 5px;
}

.vert-cent {
position: relative;
top: 50%;
transform: translateY(-50%);
}

hr
{
margin: 0;
border: 0;
height: 1.5px;
background-image: linear-gradient(to right, rgba(0, 0, 0, 0), rgba(0, 0, 0, 0.75), rgba(0, 0, 0, 0));
}
</style>

<html>
<head>
<title>Gated Shape CNN</title>
<meta property="og:title" content="gscnn" />
</head>

<body>
<br>
<center>
<span style="font-size:42px">Gated-SCNN</span>
<br>
<span style="font-size:36px">Gated Shape CNNs for Semantic Segmentation</span>
</center>

<br>
<table align=center width=700px>
<tr>
<td align=center width=100px>
<center>
<span style="font-size:20px"><a href="https://tovacinni.github.io">Towaki Takikawa</a><sup>*1,2</sup></span>
</center>
</td>

<td align=center width=100px>
<center>
<span style="font-size:20px"><a href="http://www.cs.toronto.edu/~davidj/">David Acuna</a><sup>*1,3,4</sup></span>
</center>
</td>


<td align=center width=100px>
<center>
<span style="font-size:20px"><a href="https://varunjampani.github.io/">Varun Jampani</a><sup>1</sup></span>
</center>
</td>

<td align=center width=100px>
<center>
<span style="font-size:20px"><a href="http://www.cs.toronto.edu/~fidler/">Sanja Fidler</a><sup>1,3,4</sup></span>
</center>
</td>
</tr>
</table>

<br>
<table align=center width=700px>
<tr>
<td align=center width=100px>
<center>
<span style="font-size:20px"><sup>1</sup>NVIDIA</span>
</center>
</td>
<td align=center width=100px>
<center>
<span style="font-size:20px"><sup>2</sup>University of Waterloo</span>
</center>
</td>
<td align=center width=100px>
<center>
<span style="font-size:20px"><sup>3</sup>University of Toronto</span>
</center>
</td>
<td align=center width=100px>
<center>
<span style="font-size:20px"><sup>4</sup>Vector Institute</span>
</center>
</td>
</tr>
</table>

<table align=center width=700px>
<tr>
<td align=center width=100px>
<center>
<span style="font-size:20px;color:red">ICCV, 2019</span>
</center>
</td>
</tr>
</table>
<br>
<table align=center width=900px>
<tr>
<td width=450px>
<center>
<a href="./resources/GSCNN.mp4"><img src = "./resources/gscnn.gif" width="450px" height="250px"></img>
</center>
</td>
<td width=450px>
<center>
<a href="./resources/GSCNN.mp4"><img src = "./resources/intro.jpg" width="450px" height="250px"></img></href></a><br>
</center>
</td>
<!-- -->
</tr>
</table>
<table align=center width=900px></table>
<tr>
<td width=600px>
<br>
<center>
<!-- -->
</center>
</td>
</tr>
<tr>
<td width=600px>
<br>
<p align="justify" style="font-size: 18px">

Current state-of-the-art methods for image segmentation form a dense image representation where the color, shape and texture information are all processed together inside a deep CNN. This however may not be ideal as they contain very different type of information relevant for recognition. We propose a new architecture that adds a shape stream to the classical CNN architecture. The two streams process the image in parallel, and their information gets fused in the very top layers. Key to this architecture is a new type of gates that connect the intermediate layers of the two streams. Specifically, we use the higher-level activations in the classical stream to gate the lower-level activations in the shape stream, effectively removing noise and helping the shape stream to only focus on processing the relevant boundary-related information. This enables us to use a very shallow architecture for the shape stream that operates on the image-level resolution. Our experiments show that this leads to a highly effective architecture that produces sharper predictions around object boundaries and significantly boosts performance on thinner and smaller objects. Our method achieves state-of-the-art performance on the Cityscapes benchmark, in terms of both mask (mIoU) and boundary (F-score) quality, improving by 2% and 4% over strong baselines.
</p>
</td>
</tr>
<tr>
</tr>
</table>

<br>
<hr>
<table align=center width=700>
<center><h1>News</h1></center>
<tr>
<ul>
<li>[August 2019] Code released on <a href="https://github.com/nv-tlabs/gscnn">GitHub</a></li>
<li>[July 2019] Paper accepted at ICCV 2019!</li>
<li>[July 2019] Paper released on <a href="http://arxiv.org/abs/1907.05740">arXiv</a></li>
</ul>
</tr>
</table>
<br>
<hr>
<table align=center width=700>
<center><h1>Paper</h1></center>
<tr>
<td><a href="./"><img style="height:180px; border: solid; border-radius:30px;" src="./resources/top.jpg"/></a></td>
<td><span style="font-size:18px">Towaki Takikawa* , David Acuna* , Varun Jampani , Sanja Fidler<br>
<small>(* denotes equal contribution)</small><br><br>
Gated-SCNN: Gated Shape CNNs for Semantic Segmentation<br><br>

ICCV, 2019. (to appear)<br>
</td>
</tr>
</table>
<br>

<table align=center width=700px>
<tr>
<td>
<span style="font-size:18px"><center>
<a href="http://arxiv.org/abs/1907.05740">[Preprint]</a>
</center></td>

<td><span style="font-size:18px"><center>
<a href="./resources/bibtex.txt">[Bibtex]</a>
</center></td>

<td><span style="font-size:18px"><center>
<a href="./resources/GSCNN.mp4">[Video]</a>
</center></td>

</tr>
<tr>

</tr>
</table>
<br>
<hr>

<center><h1>GSCNN in a nutshell</h1></center>
<table align=center width=1000px>
<tr>
<center>
<a href=''><img class="round" style="height:300" src="./resources/architecture.jpg"/></a>
</center>
</tr>
</table>

<br>
<hr>

<center><h1> Results</h1></center> <br>

<table align=center width=900px>
<tr>
<td width=100px>
<center>
<a href="./resources/seg.jpg"><img src = "./resources/seg.jpg" width="900px"></img></a><br>
</center>
</td>

<tr>
<td>
<center>
<span style="font-size:14px">
Qualitative Segmentation Results
</span>
</center>
</td>

</tr>
<tr>
<td colspan='2'>
<center>
<a href="./resources/edges.jpg"><img src = "./resources/edges.jpg" width="900px"></img></a><br>
</center>
</td>
</tr>

<tr>
<td colspan='2'>
<center>
<span style="font-size:14px">
Qualitative Semantic Boundary Results
</span>
</center>
</td>
</tr>

<tr>
<td colspan='2'>
<center>
<a href="./resources/table.png"><img src = "./resources/table.png" width="900px"></img></a><br>
</center>
</td>
</tr>

<tr>
<td colspan='2'>
<center>
<span style="font-size:14px">
Quantitative Results
</span>
</center>
</td>
</tr>
<tr>
<td colspan='2'>
<center>
<a href="./resources/crop.jpg"><img src = "./resources/crop.jpg" width="600px"></img></a><br>
</center>
</td>
</tr>

<tr>
<td colspan='2'>
<center>
<span style="font-size:14px">
Evaluation at different distances, measured by crop factor.
</span>
</center>
</td>
</tr>

</table>
<hr>
<br>
<table style="font-size:14px">
<tr>
<!-- -->
</table>

</body>
</html>
<!DOCTYPE html>
<meta charset="utf-8">
<title>Redirecting to https://research.nvidia.com/labs/toronto-ai/GSCNN/</title>
<meta http-equiv="refresh" content="0; URL=https://research.nvidia.com/labs/toronto-ai/GSCNN/">
<link rel="canonical" href="https://research.nvidia.com/labs/toronto-ai/GSCNN/">

0 comments on commit edd16c5

Please sign in to comment.