-
Notifications
You must be signed in to change notification settings - Fork 4
/
index.html
335 lines (278 loc) · 15 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
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<!-- ======================================================================= -->
<script src="http://www.google.com/jsapi" type="text/javascript"></script>
<script type="text/javascript">google.load("jquery", "1.3.2");</script>
<style type="text/css">
body {
font-family: "Titillium Web","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;
}
.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
{
border: 0;
height: 1px;
background-image: linear-gradient(to right, rgba(0, 0, 0, 0), rgba(0, 0, 0, 0.75), rgba(0, 0, 0, 0));
}
#authors td {
padding-bottom:5px;
padding-top:30px;
}
</style>
<!-- ======================================================================= -->
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-114291442-6"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'UA-114291442-6');
</script>
<script type="text/javascript" src="resources/hidebib.js"></script>
<link href='https://fonts.googleapis.com/css?family=Titillium+Web:400,600,400italic,600italic,300,300italic' rel='stylesheet' type='text/css'>
<head>
<meta charset="utf-8" />
<meta http-equiv="X-UA-Compatible" content="IE=edge" />
<link rel="icon" type="image/png" href="../img/favicon.ico">
<title>More Photos are All You Need: Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval</title>
<meta name="HandheldFriendly" content="True" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<script src="https://www.youtube.com/iframe_api"></script>
</head>
<body>
<table align=center width=600px>
<tr><td align=center width=600px>
<center><img src = "./sample_images/logo.png" height="75px"></img><br></center>
</td></tr>
</table>
<!-- <center><span style="font-size:44px;font-weight:bold;"></span></center><br/> -->
<center><span style="font-size:44px;font-weight:bold;">More Photos are All You Need:</span></center>
<center><span style="font-size:32px;font-weight:bold;">Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval</span></center><br/>
<table align=center width=1200px>
<tr>
<td align=center width=250px>
<center><span style="font-size:22px"><a href="http://ayankumarbhunia.github.io/" target="_blank">Ayan Kumar Bhunia</a></span></center></td>
<td align=center width=250px>
<center><span style="font-size:22px"><a href="http://www.pinakinathc.me/" target="_blank">Pinaki Nath Chowdhury</a></span></center></td>
<td align=center width=230px>
<center><span style="font-size:22px"><a href="https://aneeshan95.github.io/" target="_blank">Aneeshan Sain</a></span></center></td>
<td align=center width=200px>
<center><span style="font-size:22px"><a href="https://www.surrey.ac.uk/people/yongxin-yang/" target="_blank">Yongxin Yang</a></span></center></td>
<td align=center width=230px>
<center><span style="font-size:22px"><a href="http://personal.ee.surrey.ac.uk/Personal/T.Xiang/index.html/" target="_blank">Tao (Tony) Xiang</a></span></center></td>
<td align=center width=180px>
<center><span style="font-size:22px"><a href="http://personal.ee.surrey.ac.uk/Personal/Y.Song/" target="_blank">Yi-Zhe Song</a></span></center></td>
<tr/>
<!-- <tr>-->
<!-- <td align=center width=250px>-->
<!-- <center><span style="font-size:20px">SketchX, CVSSP, University of Surrey, United Kingdom</span></center></td>-->
<!-- <td align=center width=250px>-->
<!-- <center><span style="font-size:20px">CMU</span></center></td>-->
<!-- <td align=center width=150px>-->
<!-- <center><span style="font-size:20px">CMU/FAIR</span></center></td>-->
<!-- <td align=center width=150px>-->
<!-- <center><span style="font-size:20px">UIUC</span></center></td>-->
<!-- <tr/>-->
<!-- </table>-->
<table align=center width=1000px>
<tr>
<td align=center width=800px><center><span style="font-size:22px">SketchX, Centre for Vision Speech and Signal Processing, <br /> University of Surrey, United Kingdom </span></center></td>
<tr/>
</table><br/>
<table align=center width=700px>
<tr>
<td align=center width=700px><center><span style="font-size:22px">Published at <a href="http://cvpr2021.thecvf.com/">CVPR 2021</a> </span></center></td>
<tr/>
</table><br/>
<table align=center width=700px>
<tr>
<td align=center width=70px><center><span style="font-size:28px"><a href="https://arxiv.org/abs/2103.13990.pdf">[Paper]</a></span></center></td>
<!-- <td align=center width=70px><center><span style="font-size:28px"><a href="https://www.youtube.com/watch?v=JVyFexPEu-U">[Talk]</a></span></center></td> -->
<!-- <td align=center width=70px><center><span style="font-size:28px"><a href="https://github.com/AyanKumarBhunia/on-the-fly-FGSBIR/blob/master/images/4367-talk.pdf">[Slides]</a></span></center></td> -->
<td align=center width=100px><center><span style="font-size:28px"><a href="https://github.com/AyanKumarBhunia/semisupervised-FGSBIR">[GitHub]</a></span></center></td>
<!-- <td align=center width=100px><center><span style="font-size:28px"><a href='https://github.com/devendrachaplot/Neural-SLAM'>[GitHub Code]</a></span></center></td>-->
<tr/>
</table><br/>
<!-- <center><h2>Project Video</h2></center> -->
<table align=center width=600px>
<tr><td align=center width=600px>
<!-- <iframe width="768" height="432" src="https://youtu.be/tlyz68j_jvE" frameborder="0" allowfullscreen></iframe>-->
<!-- <iframe width="768" height="432" src="https://www.youtube.com/embed/tlyz68j_jvE" frameborder="0" allow="accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>-->
<center><img src = "./sample_images/outline.png" height="300px"></img><br></center>
</td></tr>
</table>
<br>
<div style="width:800px; margin:0 auto; text-align=justify">
A fundamental challenge faced by existing Fine-Grained Sketch-Based Image Retrieval (FG-SBIR) models is the data scarcity -- model performances are largely bottlenecked by the lack of sketch-photo pairs. Whilst the number of photos can be easily scaled, each corresponding sketch still needs to be individually produced. In this paper, we aim to mitigate such an upper-bound on sketch data, and study whether unlabelled photos alone (of which they are many) can be cultivated for performances gain. In particular, we introduce a novel semi-supervised framework for cross-modal retrieval that can additionally leverage large-scale unlabelled photos to account for data scarcity. At the centre of our semi-supervision design is a sequential photo-to-sketch generation model that aims to generate paired sketches for unlabelled photos. Importantly, we further introduce a discriminator guided mechanism to guide against unfaithful generation, together with a distillation loss based regularizer to provide tolerance against noisy training samples. Last but not least, we treat generation and retrieval as two conjugate problems, where a joint learning procedure is devised for each module to mutually benefit from each other. Extensive experiments show that our semi-supervised model yields significant performance boost over the state-of-the-art supervised alternatives, as well as existing methods that can exploit unlabelled photos for FG-SBIR.</div>
<br><hr>
<center><h1>Framework</h1></center>
<div style="width:800px; margin:0 auto; text-align=justify">
Our framework: a FG-SBIR model leverages large scale unlabelled photos using a sequential photo-to-sketch generation model along with labelled pairs. Discriminator guided instance-wise weighting and distillation loss are used to guard against the noisy generated data. Simultaneously, photo-to-sketch generation model learns by taking reward from FG-SBIR model and Discriminator via policy gradient (over both labelled and unlabelled) together with supervised VAE loss over labelled data. Note rasterization (vector to raster format) is a non-differentiable operation.
</div><br/>
<table align=center width=1000px>
<p style="margin-top:4px;"></p>
<tr><td width=1200px>
<center><a href="./sample_images/framework.png"><img src = "./sample_images/framework.png" width="800px"></img></a><br></center>
</td></tr>
</table>
<br/><hr>
<center><h1>Short Presentation</h1></center>
<table align=center width=300px>
<tr><td align=center width=300px>
<iframe width="560" height="315" src="https://www.youtube.com/embed/ZlTah9OaY9E" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture" allowfullscreen></iframe>
</td></tr>
</table>
<br><hr>
<!--https://youtu.be/JVyFexPEu-U-->
<table align=center width=850px>
<center><h1>Bibtex</h1></center>
<tr>
<td width=200px align=left>
<!-- <p style="margin-top:4px;"></p> -->
<!-- <a href="https://arxiv.org/pdf/2002.10310.pdf"><img style="width:200px" src="./images/4367-teaser.gif"/></a> -->
<center>
<!-- <span style="font-size:20pt"><a href="https://arxiv.org/pdf/2002.10310.pdf">[Paper]</a> -->
<!-- <span style="font-size:20pt"><a href="https://arxiv.org/abs/1902.05546v2">[ArXiv]</a>-->
<!-- <span style="font-size:20pt"><a href="resources/slides.pdf">[Slides]</a></span>-->
<!-- <span style="font-size:20pt"><a href="resources/poster.pdf">[Poster]</a></span>-->
</center>
</td>
<td width=50px align=center>
</td>
<td width=550px align=left>
<!-- <p style="margin-top:4px;"></p> -->
<!-- Chaplot, D.S., Gandhi, D., Gupta, S., Gupta, A. and Salakhutdinov, R., 2020. Learning To Explore Using Active Neural SLAM. In International Conference on Learning Representations (ICLR).-->
<p style="text-align:left;"><b><span style="font-size:20pt">Citation</span></b><br/><span style="font-size:6px;"> <br/></span> <span style="font-size:15pt"> More Photos are All You Need: Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval. In CVPR 2021.</span></p>
<!-- <p style="margin-top:20px;"></p> -->
<span style="font-size:20pt"><a shape="rect" href="javascript:togglebib('assemblies19_bib')" class="togglebib">[Bibtex]</a></span>
<div class="paper" id="assemblies19_bib">
<pre xml:space="preserve">
@InProceedings{bhunia_semifgsbir,
author = {Ayan Kumar Bhunia and Pinaki Nath Chowdhury and Aneeshan Sain and Yongxin Yang and Tao Xiang and Yi-Zhe Song},
title = {More Photos are All You Need: Semi-Supervised Learning for Fine-Grained Sketch Based Image Retrieval},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2021}
}
</pre>
</div>
</td>
</tr>
<tr>
<td width=250px align=left>
</td>
<td width=50px align=center>
</td>
<td width=550px align=left>
</td>
</tr>
</table>
<!-- <br><hr>
<div style="width:800px; margin:0 auto; text-align=justify">
</div><br/>
<table align=center width=1000px>
<p style="margin-top:4px;"></p>
<tr><td width=1200px>
<center><a href="./images/4367-teaser.gif"><img src = "./images/4367-teaser.gif" width="512px"></img></a><br></center>
</td></tr>
</table>
<br/><hr>
<table align=center width=800px>
<tr><td width=800px><left> -->
<!-- <center><h1>Acknowledgements</h1></center>
<br>
Website template from <a href="https://richzhang.github.io/colorization">here</a> and <a href="https://pathak22.github.io/modular-assemblies/">here</a>. <br>
</left></td></tr>
</table>
<br><br>
<script xml:space="preserve" language="JavaScript">
<table align=center width=300px>
<tr><td align=center width=300px>
<script type="text/javascript" id="clustrmaps" src="//clustrmaps.com/map_v2.js?d=F5sv0Hja_rXL_5biv38usYAUtI6axFYvv-8QX8EwtSk&cl=ffffff&w=a"></script>
</td></tr>
</table> -->
<!-- hideallbibs(); -->
</script>
</body>
</html>