-
Notifications
You must be signed in to change notification settings - Fork 2
/
PFSNet.html
executable file
·146 lines (144 loc) · 6.17 KB
/
PFSNet.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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>PFSNet</title>
<link rel="stylesheet" type="text/css" href="assets/scripts/bulma.min.css">
<link rel="stylesheet" type="text/css" href="assets/scripts/theme.css">
<link rel="stylesheet" type="text/css" href="https://cdn.bootcdn.net/ajax/libs/font-awesome/4.7.0/css/font-awesome.min.css">
</head>
<body>
<section class="hero is-light" style="">
<div class="hero-body" style="padding-top: 50px;">
<div class="container" style="text-align: center;margin-bottom:5px;">
<h1 class="title">
Pyramidal Feature Shrinking for Salient Object Detection
</h1>
<div class="author">Mingcan Ma<sup>1,2</sup></div>
<div class="author">Changqun Xia<sup>2</sup></div>
<div class="author">Jia Li<sup>1,2</sup></div>
<div class="group">
<a href="http://cvteam.net/">CVTEAM</a>
</div>
<div class="aff">
<p><sup>1</sup>State Key Laboratory of Virtual Reality Technology and Systems, SCSE, Beihang University, Beijing, China</p>
<p><sup>2</sup>Pengcheng Laboratory, Shenzhen, China</p>
</div>
<div class="con">
<p style="font-size: 24px; margin-top:5px; margin-bottom: 15px;">
AAAI 2021
</p>
</div>
<div class="columns">
<div class="column"></div>
<div class="column"></div>
<div class="column">
<a href="https://ojs.aaai.org/index.php/AAAI/article/view/16331" target="_blank">
<p class="link">Paper</p>
</a>
</div>
<div class="column">
<a href="https://github.com/iCVTEAM/PFSNet/" target="_blank">
<p class="link">Code</p>
</a>
</div>
<div class="column"></div>
<div class="column"></div>
</div>
</div>
</div>
</section>
<div style="text-align: center;">
<div class="container" style="max-width:850px">
<div style="text-align: center;">
<img src="assets/PFSNet/head.png" class="centerImage">
</div>
</div>
<div class="head_cap">
<p style="color:gray;">
The overall display of PFSNet
</p>
</div>
</div>
<section class="hero">
<div class="hero-body">
<div class="container" style="max-width: 800px" >
<h1 style="">Abstract</h1>
<p style="text-align: justify; font-size: 17px;">
Recently, we have witnessed the great progress of salient object
detection (SOD), which benefits from the effectiveness
of various feature aggregation strategies. However, existing
methods usually aggregate the low-level features containing
details and the high-level features containing semantics over
a large span, which introduces noise into the aggregated features
and generates inaccurate saliency maps. In this paper,
we propose a pyramidal feature shrinking network (PFSNet),
which aims to aggregate adjacent feature nodes in pairs with
layer-by-layer shrinkage, so that the aggregated features fuse
effective details and semantics and discard interference information.
Specifically, a pyramidal shrinking decoder (PSD) is
proposed to aggregate adjacent features hierarchically in an
asymptotic manner. Unlike other methods that aggregate features
with significantly different information, this method only
focuses on adjacent feature nodes in each layer and shrinks
them to a final unique feature node. Besides, we propose
an adjacent fusion module (AFM) to perform mutual spatial
enhancement between the adjacent features to dynamically
weight the features and adaptively fuse the appropriate information.
Besides, a scale-aware enrichment module (SEM)
based on the features extracted from the backbone is utilized
to obtain rich scale information and generate diverse initial
features with dilated convolutions. Extensive quantitative and
qualitative experiments demonstrate that the proposed intuitive
framework outperforms 14 state-of-the-art approaches
on 5 public datasets.
</p>
</div>
</div>
</section>
<section class="hero is-light" style="background-color:#FFFFFF;">
<div class="hero-body">
<div class="container" style="max-width:800px;margin-bottom:20px;">
<h1>
Qualitative Comparison
</h1>
</div>
<div class="container" style="max-width:800px">
<div style="text-align: center;">
<img src="assets/PFSNet/comp.png" class="centerImage">
</div>
</div>
</div>
</section>
<section class="hero" style="padding-top:0px;">
<div class="hero-body">
<div class="container" style="max-width:800px;">
<div class="card">
<header class="card-header">
<p class="card-header-title">
BibTex Citation
</p>
<a class="card-header-icon button-clipboard" style="border:0px; background: inherit;" data-clipboard-target="#bibtex-info" >
<i class="fa fa-copy" height="20px"></i>
</a>
</header>
<div class="card-content">
<pre style="background-color:inherit;padding: 0px;" id="bibtex-info">@article{Ma_Xia_Li_2021,
title={Pyramidal Feature Shrinking for Salient Object Detection},
volume={35},
url={https://ojs.aaai.org/index.php/AAAI/article/view/16331},
number={3},
journal={Proceedings of the AAAI Conference on Artificial Intelligence},
author={Ma, Mingcan and Xia, Changqun and Li, Jia},
year={2021},
month={May},
pages={2311-2318}
}</pre>
</div>
</section>
<script type="text/javascript" src="assets/scripts/clipboard.min.js"></script>
<script>
new ClipboardJS('.button-clipboard');
</script>
</body>
</html>