forked from shijieS/DMMN_Page
-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathindex.html
288 lines (257 loc) · 9.59 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
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
<!DOCTYPE html>
<!-- modified from url=(0049)https://github.com/shijieS/DMMN_Page -->
<html class="gr__ee_nthu_edu" lang="en">
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
<meta content="IE=edge" http-equiv="X-UA-Compatible">
<meta content="width=device-width, initial-scale=1" name="viewport">
<meta content="XueFei" name="author">
<title>An Active Terahertz Imaging Dataset for Concealed Multi-object Detection and Evaluation</title>
<!-- CSS includes -->
<link href="./asset/bootstrap.min.css" rel="stylesheet">
<link href="./asset/css" rel="stylesheet" type="text/css">
<link href="./asset/mystyle.css" rel="stylesheet">
</head>
<body data-gr-c-s-loaded="true">
<div class="topnav" id="myTopnav">
<a href="#header">Home</a>
<a href="#abstract">Abstract</a>
<a href="#dataset">Dataset</a>
<a href="#paper">Paper</a>
<a href="#acknowledgement">Acknowledgements</a>
<a class="icon" href="javascript:void(0);" onclick="toggleTopNav()">☰</a>
</div>
<div class="container-fluid" id="header">
<div class="row">
<h1>An Active Terahertz Imaging Dataset for Concealed Multi-object Detection and Evaluation</h1>
<div class="authors">
<a href="https://scholar.google.com/citations?user=oKE2Wx8AAAAJ&hl=en&oi=ao" target="_blank">Dong Liang,</a>
Fei Xue
<br><br>
<p style="text-align:center;">
<a href="https://www.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa.jpg">
</a>
<a href="http://cs.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa-ccst.png">
</a>
<a href="http://cs.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa-ai.png">
</a>
<a href="https://soft2011.nju.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/cicnsti.png">
</a>
</p>
</div>
</div>
</div>
<div class="container" id="abstract">
<h2>Abstract</h2>
<p style="text-align:justify;">
Concealed objects detection in Terahertz imaging is an urgent need for public security and counter-terrorism.
So far, there is no public Terahertz imaging dataset for evaluation of objects detection algorithms. This paper
provides a public dataset for evaluating multi-object detection algorithms in active Terahertz imaging worked at
140 GHz, with the imaging resolution 5 mm by 5 mm. Due to poor imaging quality, object detection on this dataset
is much more difficult than on those commonly used public object detection datasets in computer vision field.
Several state-of-the-art computer vision based detectors, including YOLOv3, YOLOv4, FRCN-OHEM and RetinaNet, are
evaluated on this dataset. In addition, we enhance RetinaNet through embedding low-level features for detecting
small objects. Aiming at solving the problem of unbalanced and hard training samples, Focal Loss and Online Hard
Example Mining technology are discussed and employed. Experimental results show that the enhanced RetinaNet
achieves the best mAP with fast detection speed which can meet the requirement of real-time security inspection.
Experiment also indicates that hiding objects in different parts of the human body affect the detection accuracy.
</p>
</div>
<div class="container" id="dataset">
<h2>Dataset</h2>
<p style="text-align:center;">
<img width="100%" src="./ICASSP2021/classes.png">
<br>Sample display for each category. The bottom row shows a zoomed-in view of the object.
</p><br><br>
<p style="text-align:center;">
<img width="100%" src="./ICASSP2021/diversification.png">
<br>Sample display for diversification.
</p>
<br>
<h3>Dataset information:</h3>
Category table:<br>
<table style="width:100%;" border=1>
<tr>
<th>Class</th>
<td>GA</td>
<td>KK</td>
<td>SS</td>
<td>MD</td>
<td>CK</td>
<td>WB</td>
<td>KC</td>
<td>CP</td>
<td>CL</td>
<td>LW</td>
<td>UN</td>
</tr>
<tr>
<th>Item</th>
<td>Gun</td>
<td>Kitchen Knife</td>
<td>Scissors</td>
<td>Metal Dagger</td>
<td>Ceramic Knife</td>
<td>Water Bottle</td>
<td>Key Chain</td>
<td>Cell Phone</td>
<td>Cigarette Lighter</td>
<td>Leather Wallet</td>
<td>Unknown</td>
</tr>
<tr>
<th>Quantity</th>
<td>116</td>
<td>100</td>
<td>96</td>
<td>64</td>
<td>129</td>
<td>107</td>
<td>78</td>
<td>129</td>
<td>163</td>
<td>78</td>
<td>289</td>
</tr>
</table>
<br>
Dataset format table:
<table style="width:100%;" border=1>
<tr>
<th>Number of images</th>
<th>Image size and format</th>
<th>Imaging resolution</th>
<th>Models</th>
<th>Number of categories</th>
<th>Objects per image</th>
<th>Maximum object size</th>
<th>Average object size</th>
<th>Minimum object size</th>
</tr>
<tr>
<td>3157</td>
<td>335×880 p.x. JPEG</td>
<td>5×5 mm</td>
<td>4 males, 6 females</td>
<td>11</td>
<td>0,1,2,3</td>
<td>13390 p.x.</td>
<td>3222 p.x.</td>
<td>390 p.x.</td>
</tr>
</table>
<h3>Dataset structure:</h3>
THZ_dataset_det_VOC<br>
  ├── Annotations<br>
  │    ├── D_N_F1_CK_F_LA_WB_F_S_back_0907140917.xml<br>
  │    ├── D_N_F1_CK_F_LA_WB_F_S_front_0907140917.xml<br>
  │    ├── D_N_F1_CL_V_LA_LW_V_RA_back_0907141138.xml<br>
  │    ├── ...<br>
  │    └── T_P_M6_MD_F_LL_CK_F_C_WB_F_RT_front_0906154134.xml<br>
  └── JPEGImages<br>
      ├── D_N_F1_CK_F_LA_WB_F_S_back_0907140917.jpg<br>
      ├── D_N_F1_CK_F_LA_WB_F_S_front_0907140917.jpg<br>
      ├── D_N_F1_CL_V_LA_LW_V_RA_back_0907141138.jpg<br>
      ├── ...<br>
      └── T_P_M6_MD_F_LL_CK_F_C_WB_F_RT_front_0906154134.jpg<br>
<br>
<h3>Download links:</h3>
<div class="row">
<div class="col-md-4">
<a href="https://drive.google.com/file/d/1YVAkhoYjSCg3r5eSM2s7zhBSLEWv3YB-/view?usp=sharing" target="_blank">
<p style="text-align:center;">
<img width="100" src="./ICASSP2021/google-cloud.jpg"><br>
Google drive<br>
</a>
</div>
<div class="col-md-4">
<a href="https://pan.baidu.com/s/1Bfn8ra4LKmUE0qkR0sCwZA" target="_blank">
<p style="text-align:center;">
<img width="100" src="./ICASSP2021/baidu-cloud.jpg"><br>
Baidu drive<br>
</a>
(Extraction code: yuax)
</div>
<div class="col-md-4">
<a href="https://pan.nuaa.edu.cn/share/6cd1ccc7447c5a05dd61782f31" target="_blank">
<p style="text-align:center;">
<img width="100" src="./ICASSP2021/nuaa.jpg"><br>
NUAA drive<br>
</a>
</div>
</div>
</div>
<div class="container" id="paper">
<h2>Submitted to ICASSP2021</h2>
<!--
<div>
<pre class="citation">@inproceedings{DLiang2021,
author = {Dong Liang, Fei Xue},
title = {An Active Terahertz Imaging Dataset for Concealed Multi-object Detection and Evaluation},
booktitle = {International Conference on Acoustics, Speech and Signal Processing (ICASSP2021)},
year = {2021}
}</pre>
</div>
-->
</div>
<div class="container" id="acknowledgement">
<h2>Acknowledgements</h2>
<p style="text-align:center;">
<a href="https://www.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa.jpg">
</a>
<a href="http://cs.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa-ccst.png">
</a>
<a href="http://cs.nuaa.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/nuaa-ai.png">
</a>
<a href="https://soft2011.nju.edu.cn/" target="_blank">
<img height="100" src="./ICASSP2021/cicnsti.png">
</a>
<a href="https://www.caep.cn/" target="_blank">
<img height="100" src="./ICASSP2021/caep.jpg">
</a>
<a href="https://www.nvidia.com" target="_blank">
<img height="100" src="./ICASSP2021/nvidia.jpg">
</a>
<a href="https://www.tensorflow.org/" target="_blank">
<img height="100" src="./ICASSP2021/tensorflow.png">
</a>
</p>
</div>
<div id="footer">
<br>
<p style="text-align:center;">
Copyright © Fei Xue 2021 Based on
<a href="https://github.com/shijieS/DMMN_Page" target="_blank">DMMN</a>
<br>Last modified on January 13, 2021
</p>
</div>
<!-- Javascript includes -->
<script src="./asset/jquery-1.8.3.min.js"></script>
<script src="./asset/mystyle.js"></script>
<script src="./asset/bootstrap.min.js"></script>
<script async="" src="./asset/analytics.js"></script>
<script>
(function (i, s, o, g, r, a, m) {
i['GoogleAnalyticsObject'] = r;
i[r] = i[r] || function () {
(i[r].q = i[r].q || []).push(arguments)
}, i[r].l = 1 * new Date();
a = s.createElement(o),
m = s.getElementsByTagName(o)[0];
a.async = 1;
a.src = g;
m.parentNode.insertBefore(a, m)
})(window, document, 'script', '//www.google-analytics.com/analytics.js', 'ga');
ga('create', 'UA-98479202-1', 'auto');
ga('send', 'pageview');
</script>
<div id="point-jawn" style="z-index: 2147483647;"></div>
</body>
</html>