-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
225 lines (213 loc) · 9.51 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
<!DOCTYPE html>
<html lang="en-us">
<head>
<meta charset="UTF-8">
<title>AirBirds Dataset</title>
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" type="text/css" href="stylesheets/normalize.css" media="screen">
<link href='https://fonts.googleapis.com/css?family=Open+Sans:400,700' rel='stylesheet' type='text/css'>
<link rel="stylesheet" type="text/css" href="stylesheets/stylesheet.css" media="screen">
<link rel="stylesheet" type="text/css" href="stylesheets/github-light.css" media="screen">
<style>
table {
width: 100%;
}
.styled-table {
border-collapse: collapse;
font-size: 0.9em;
font-family: sans-serif;
min-width: 400px;
box-shadow: 0 0 20px rgba(0, 0, 0, 0.15);
}
.styled-table thead tr {
background-color: #009879;
color: #ffffff;
text-align: left;
}
.styled-table th,
.styled-table td {
width:12.5%;
padding: 12px 15px;
}
.styled-table tbody tr {
border-bottom: 1px solid #dddddd;
}
.styled-table tbody tr:nth-of-type(even) {
background-color: #f3f3f3;
}
.styled-table tbody tr:last-of-type {
border-bottom: 2px solid #009879;
}
.styled-table tbody tr.active-row {
font-weight: bold;
color: #009879;
}
</style>
</head>
<body>
<section class="page-header" style="backgroundImage:url('images/background1.png');background-size: 100% 100%;">
<h1 class="project-name">AirBirds<h1>
<h2 class="project-tagline">A Large-scale Challenging Dataset for Bird Strike Prevention in Real-world Airports
</h2>
</section>
<section class="main-content">
<h1>
<a id="listen-1" class="anchor" href="#listen-1" aria-hidden="true"><span aria-hidden="true"
class="octicon octicon-link"></span></a>Introduction
</h1>
<p style="text-align: justify;text-indent: 2ch;">One fundamental limitation to the research of bird strike
prevention is the lack of a large-scale dataset taken
directly from real-world airports. Existing relevant datasets are either small in size or not dedicated for this
purpose. To advance the research and practical solutions for bird strike prevention, we present a
large-scale challenging dataset AirBirds that consists of 118,312 time-series images, where a total of 409,967
bounding boxes of flying birds are manually, carefully annotated. The average size of all annotated instances is
smaller than 10 pixels in 1920x1080 images. Images in the dataset are captured over 4 seasons of a whole
year by a network of cameras deployed at a real-world airport, covering diverse bird species, lighting conditions
and weather scenarios. To the best of our knowledge, it is the first large-scale image dataset that collects
flying birds in real-world airports for bird strike prevention. Since the detection module is a crucial component
of the strike prevention system, we also investigate the performances of existing representative detectors on
AirBirds. The results show that their performances are surprisingly unsatisfactory and the reasons are analyzed in
detail. Due to special data distribution, AirBirds can also serve as a challenging benchmark for tiny object
detection.</p>
<p align="center"> <iframe width="830" height="450" src="https://www.youtube.com/embed/TSS_bOM8uSY" frameborder="0"
allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen
style="max-width: 100%"></iframe>
</p>
<p style="text-align: justify;text-indent: 2ch;">
In addition, we develop a vision-based system to detect flying birds in a real-world airport. The above video
demonstrates this case.
Subsequent measures of bird repelling are triggered automatically once targets are discovered in the specified areas.
</p>
<h2>
<a id="download" class="anchor" href="#%E5%AE%89%E8%A3%85" aria-hidden="true"><span aria-hidden="true"
class="octicon octicon-link"></span></a>Download
</h2>
<p style="text-align: justify;text-indent: 2ch;">
The totoal size of all compressed files in AirBirds is about 165GB, it is necessary to split the dataset into multiple chunks,
as the following table summarizes. The training set includes images0.zip, ..., images9.zip and corresponding labels0.zip, ..., labels9.zip,
remaining images10.zip and images11.zip are in the test set. The dataset usage instructions and download links can be found
by filling out this <a href="https://forms.gle/BCydUZpQMm9ZQQMR9">FORM</a>,
including the <a href="https://pan.baidu.com/s/1iDhagXsSsMAh9l33Mv93Tw">Baidu NetDisk</a> extraction code
and another download link when you cannot access Baidu NetDisk.
</p>
<table class="styled-table">
<thead>
<tr>
<th>Chunk</th>
<th>Images</th>
<th>Labels</th>
<th>Size</th>
<th>Chunk</th>
<th>Images</th>
<th>Labels</th>
<th>Size</th>
</tr>
</thead>
<tbody>
<tr>
<td>0</td>
<td>images0.zip</td>
<td>labels0.zip</td>
<td>13GB</td>
<td>1</td>
<td>images1.zip</td>
<td>labels1.zip</td>
<td>13GB</td>
</tr>
<tr>
<td>2</td>
<td>images2.zip</td>
<td>labels2.zip</td>
<td>13GB</td>
<td>3</td>
<td>images3.zip</td>
<td>labels3.zip</td>
<td>12GB</td>
</tr>
<tr>
<td>4</td>
<td>images4.zip</td>
<td>labels4.zip</td>
<td>14GB</td>
<td>5</td>
<td>images5.zip</td>
<td>labels5.zip</td>
<td>13GB</td>
</tr>
<tr>
<td>6</td>
<td>images6.zip</td>
<td>labels6.zip</td>
<td>14GB</td>
<td>7</td>
<td>images7.zip</td>
<td>labels7.zip</td>
<td>16GB</td>
</tr>
<tr>
<td>8</td>
<td>images8.zip</td>
<td>labels8.zip</td>
<td>21GB</td>
<td>9</td>
<td>images9.zip</td>
<td>labels9.zip</td>
<td>11GB</td>
</tr>
<tr>
<td>10</td>
<td>images10.zip</td>
<td>---------</td>
<td>16GB</td>
<td>11</td>
<td>images11.zip</td>
<td>---------</td>
<td>15GB</td>
</tr>
</tbody>
</table>
<h2>
<a id="gallery" class="anchor" href="#%E5%AE%89%E8%A3%85" aria-hidden="true"><span aria-hidden="true"
class="octicon octicon-link"></span></a>Gallery
</h2>
<p style="text-align: justify;text-indent: 2ch;">
To provide you an intuitive impression of AirBirds, a gallery that consists of 64 random samples is presented.
Samples exhibited here are only a small portion
of 118,312 images and we hope they open a window of exploring this challenging and scenario-diverse dataset for the research of bird
strike prevention.
</p>
<p align="center"><img src="images/gallery.jpg" height="450" width="830" />
</p>
<h2>
<a id="citation" class="anchor" href="#citation" aria-hidden="true"><span aria-hidden="true"
class="octicon octicon-link"></span></a>Citation
</h2>
<p style="text-align: justify;text-indent: 2ch;">
If you use our dataset or find our dataset is helpful for your research and project, please cite our paper as follows <br>
@InProceedings{Sun_2022_ACCV, <br>
author = {Sun, Hongyu and Wang, Yongcai and Cai, Xudong and Wang, Peng and Huang, Zhe and Li, <br>
Deying and Shao, Yu and Wang, Shuo}, <br>
title = {AirBirds: A Large-scale Challenging Dataset for Bird Strike Prevention in Real-world Airports}, <br>
booktitle = {Proceedings of the Asian Conference on Computer Vision (ACCV)}, <br>
month = {December}, <br>
year = {2022}, <br>
pages = {2440-2456} <br>
}
</p>
<h2>
<a id="license" class="anchor" href="#license" aria-hidden="true"><span aria-hidden="true"
class="octicon octicon-link"></span></a>License
</h2>
<p style="text-align: justify;text-indent: 2ch;">AirBirds is for non-commercial research and educational use only. Researchers can share and adapt this dataset under the license
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International
(<a href="https://creativecommons.org/licenses/by-nc-sa/4.0/">CC BY-NC-SA 4.0</a>).
</p>
<br>
<script type="text/javascript" id="clustrmaps" src="//clustrmaps.com/map_v2.js?d=00GNRNfgX-cD0jIwv3h2voCkNi2wR31F6OmY6OhGXMw&cl=ffffff&w=a"></script>
<footer class="site-footer" align="center">
<span class="site-footer-owner">This website is maintained by <a
href="https://airbirdsdata.github.io/">AirBirds Team</a>.</span>
</footer>
</section>
</body>
</html>