-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathDatasets.html
More file actions
123 lines (110 loc) · 5.78 KB
/
Copy pathDatasets.html
File metadata and controls
123 lines (110 loc) · 5.78 KB
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
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=windows-1252"
http-equiv="Content-Type">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="shortcut icon" href="1.jpg">
<link rel="stylesheet" type="text/css" href="style2.css" />
<title>Liang Zheng</title>
</head>
<body>
<div class="main-container">
<style>
a{ text-decoration: none;}
</style>
<div id="header">
<table style="text-align: left; width: 100%;" cellspacing="2"
cellpadding="2" border="0">
<tbody>
<tr>
<td colspan="3" rowspan="1" style="vertical-align: top;"><button type="button" class="nav-toggle" aria-label="Menu" onclick="this.parentNode.querySelector('.nav-links').classList.toggle('open')">☰</button><span class="nav-links"
style="font-family: Arial; font-size: 22px">
<a style="text-decoration:underline;"
href="index.html"><font color=black>Home</font></a>
<a style="text-decoration:underline;"
href="people.html"><font color=black>People</font></a>
<a style="text-decoration:underline;"
href="Publication.html"><font color=black>Publications</font></a>
<a style="text-decoration:underline;"
href="Code.html"><font color=black>Code</font></a>
<a style="text-decoration:underline;"
href="Datasets.html"><font color=black>Datasets</font></a>
<a style="text-decoration:underline;"
href="Slides.html"><font color=black>Slides</font></a>
<br><br>
</table>
</div>
<div id="papers">
<br /><font size=4.0 color="black"><strong>Alice Benchmarks</strong></font>
<br /><span style="line-height:22px;">
Alice is an online evaluation server that contains a series of domain adaptation tasks in computer vision. The source is simulated data; the target is real or synthetic data.
<br />
<a href="http://45.32.72.229/"><font color="blue">project page</font></a>
<br />
<a href="http://45.32.72.229/alice/challenge-1/"><font color="blue">domain adaptive pedestrian recognition</font></a>
<br />
More tasks are coming, aiming at content-level domain adaptation. That is, the source domain is editable through graphics engines.
<br /><br />
<div id="papers">
<br /><font size=4.0 color="black"><strong>PersonX engine</strong></font>
<br /><span style="line-height:22px;">
This is a data synthesis engine capable of generating persons under controllable
environmental conditions. That is, the illumination, viewpoint, camera settings can be edited precisely. In our CVPR19 paper, it is used to assess the impact of environmental factors on the system accuracy.
<br />
<a href="https://github.com/sxzrt/Dissecting-Person-Re-ID-from-the-Viewpoint-of-Viewpoint"><font color="blue">project page</font></a>
<br /><br />
<div id="papers">
<br /><font size=4.0 color="black"><strong>Market-1501 dataset</strong></font>
<br /><span style="line-height:22px;">
This dataset is collected and annotated in Tsinghua. It has 6 cameras, 1,501 IDs and a distractor set of 500k images. Only one train/test split is used. We also annotate the ID-level attributes for Market-1501.
<br />
<a href="Project/project_reid.html"><font color="blue">project page</font></a>
<br />
<a href="https://github.com/vana77/Market-1501_Attribute"><font color="blue">attributes</font></a>
<br />
<a href="./Project/state_of_the_art_market1501.html"><font color="blue">state of the art on Market-1501</font></a>
<br />
<a href="Project/state_of_the_art_500k.html"><font color="blue">state of the art on Market-1501+500k</font></a>
<br /><br />
<font size=4.0 color="black"><strong>CUHK03 dataset with new evaluation protocol</strong></font>
<br />
We have used a new evaluation protocol for the CUHK03 dataset. This protocol only has one train/test split instead of the previous 20 train/test splits. The training and testing sets have 767 and 700 IDs, respectively.
<br />
<a href="https://github.com/zhunzhong07/person-re-ranking"><font color="blue">project page</font></a>
<br /><br />
<font size=4.0 color="black"><strong>MARS dataset</strong></font>
<br />
This dataset is a video extension of Market-1501. It has 6 cameras and 1,261 IDs. Only one train/test split is used.
<br />
<a href="Project/project_mars.html"><font color="blue">project page</font></a>
<br />
<a href="Project/state_of_the_art_mars.html"><font color="blue">state of the art on MARS</font></a>
<br /></font></a>
<br />
<font size=4.0 color="black"><strong>PRW dataset</strong></font>
<br />
This dataset is another extension of Market-1501. It provides 11,816 video frames. We draw a bounding box for all the pedestrians. We assign an ID (1-932) to most of them, and -2 to those whose IDs we do not know for sure. For this dataset, one should create a gallery by pedestrian detection and then evaluate person re-identification. Re-ID performance is closely related to detection performance.
<br />
<a href="Project/project_prw.html"><font color="blue">project page</font></a>
<br />
</span>
</div>
<!--</div>-->
<br class="clearfix" />
<!--</div>-->
<br class="clearfix" />
</div>
</div>
<!-- <div id="footer">
Last Update: Apr. 25, 2013 by Liang Zheng.
</div> -->
</div>
</body>
</html>