-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
165 lines (140 loc) · 11.3 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
<!DOCTYPE html>
<!-- -->
<html>
<body>
<center><h2> Beyond Pixels: Leveraging Geometry and Shape Cues for Online Multi-Object Tracking </h2></center>
<center><h4><i> Sarthak Sharma<sup>*</sup>, Junaid Ahmed Ansari<sup>*</sup>, J. Krishna Murthy, and K. Madhava Krishna</i></h4></center>
<center><h4>International Institute of Information Technology, Hyderabad. INDIA </h4></center>
<center><font size="2"> ( *Authors with equal contribution)</font></center>
<!-- <center><font size="1">(The first two authors contributed equally to this work)</font></center><br> -->
<center><img src="teaser_final.png" alt="Mountain View" style="width:760px;height:280px;">
<iframe width="500" height="280" src="https://www.youtube.com/embed/DpmzQSXhZko?rel=0" frameborder="0" allowfullscreen></iframe>
</center><br>
<center><font size="1.5">(Above figure: An illustration of the proposed method. The first two rows show objects tracks in frames t and t + 1. The bottom row depicts how 3D position
and orientation information is propagated from frame t to frame t + 1. This information is used to specify search areas for each object in the subsequent
frame, and this greatly r educes the number of pairwise costs that are to be computed.) </font></center>
<h3> Code can be found <a href="https://github.com/JunaidCS032/MOTBeyondPixels">here</a>.</h3>
<h2>ABSTRACT</h2>
This paper introduces geometry and novel object shape and pose costs for multi-object tracking in road
scenes. Using images from a monocular camera alone, we devise pairwise costs for object tracks, based on several 3D
cues such as object pose, shape, and motion. The proposed costs are agnostic to the data association method and can be
incorporated into any optimization framework to output the pairwise data associations. These costs are easy to implement,
can be computed in real-time, and complement each other to account for possible errors in a tracking-by-detection frame-
work. We perform an extensive analysis of the designed costs and empirically demonstrate consistent improvement over the
state-of-the-art under varying conditions that employ a range of object detectors, exhibit a variety in camera and object
motions, and, more importantly, are not reliant on the choice of the association framework. We also show that, by using
the simplest of associations frameworks (two-frame Hungarian assignment), we surpass the state-of-the-art in multi-object-
tracking on road scenes.
<h2>COMPARISION WITH PREVIOUS PUBLISHED APPROACHES</h2>
<h4><font color="green"> The proposed method (MOTBeyondPixels) is currently third (it was 1st amongs the published approaches on the time of sumbission)
on the KITTI Object Tracking leaderboard. Evaluation results can be found <a href="http://www.cvlibs.net/datasets/kitti/eval_tracking_detail.php?result=43713ff142271660c77d258af6bdbceb079db678">here</a>.</font></h4>
<center>
<img src="MOTA.png" alt="Mountain View" style="width:500px;height:430px;">
<img src="kitti_bench_best.png" alt="Mountain View" style="width:450px;height:250px;">
</center><br>
<center><font size="2">Left: Our method outperforms all published multi-object trackers. Right: Snapshot of KITTI Tracking leaderboard where our method is shown as MOTBeyondPixels</font></center><br>
<font size="2"><b>MOTA(Multi-Object Tracking Accuracy):</b> Indicates how consistently a target is tracked.</font><br>
<font size="2"><b>MOTP(Multi-Object Tracking Precision):</b> Indicates how precisely target is localized.</font><br>
<font size="2"><b>MT(Mostly Tracked):</b> Percentage of cars that are tracked correctly for 80% of the trajectory. Higher is better</font><br>
<font size="2"><b>ML(Mostly Lost):</b> Percentage of cars that are not correctly tracked for 80% of the trajectory. Lower is better</font><br>
<font size="2">(Please note that our method is completely online i.e. two frame based approach, and no optimization is applied. We will soon be releasing the code with
and without optimization.)</font>
<h2>TRACKING RESULTS ON KITTI TEST SEQUENCES</h2>
<!-- seq 1,2 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/zEENcR_FGUg?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/Jw54e_Irw3M?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 3,4 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/u3IERzjb6uM?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/q-9AJkHgvOw?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 5,6 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/_vQHvgB3fiA?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/o4_P0PYX4Dk?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 7,8 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/zWatqMPYMnc?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/GVLB6YHvvUs?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 9,10 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/i08p_Kn2sds?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/J1aC4KYAV0A?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 11,12 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/CcKmDPMTtdk?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/1uS9kyEa_cQ?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 13,14 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/3HwrYzxKckc?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/gfKKp_y-0W8?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 15,16 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/v-R_MUqiMlA?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/3JVD5ywId4I?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 17,18 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/koF9QdbtMY8?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/K8mYRTft2qo?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 19,20 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/lxA11aWPLi8?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/LpM0o0lvl-Y?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 21,22 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/GO4xjjHZPqU?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/IQv43zGWG-Y?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 23,24 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/VBoMHh9Sdhs?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/rB4ME36FTjM?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 25,26 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/7fWZqj8RJQM?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/wFLH623sXFs?rel=0" frameborder="0" allowfullscreen></iframe>
<!-- seq 27,28 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/HNP0ceTOWuo?rel=0" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/HNP0ceTOWuo?rel=0" frameborder="0" allowfullscreen></iframe>
<h2>TRACKING RESULTS ON KITTI TRAIN SEQUENCES</h2>
<!-- seq 1,2 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/jDytZOI3aCM?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/DOlbuxWvhpE?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 3,4 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/FwcbMwrNqvU?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/lo8kLFsPy7Y?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 5,6 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/0iJJBNpYptw?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/qn0CcJfogWE?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 7,8 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/4Awdd9tRw6g?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/anlpTpmZI3o?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 9,10 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/WCU2hmBaLpk?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/iGNkB0rLS_E?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 11,12 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/FzCT_ami5zg?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/0wGwXU4_L5A?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 13,14 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/S41rBGl1clg?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/xCD4G30iCAM?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 15,16 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/JN6bsh4nwxc?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/ZAVXw5stUyo?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 17,18 -->
<br>
<iframe width="560" height="315" src="https://www.youtube.com/embed/f3kd9-NlKrA?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/6u9QDCXkcew?ecver=1" frameborder="0" allowfullscreen></iframe>
<!-- seq 19,20 -->
<iframe width="560" height="315" src="https://www.youtube.com/embed/U-q_Njx6jec?ecver=1" frameborder="0" allowfullscreen></iframe>
<iframe width="560" height="315" src="https://www.youtube.com/embed/34evFQC_wg0?ecver=1" frameborder="0" allowfullscreen></iframe>
</body>
</html>