Method

MOTBeyondPixels [on] [MOTBeyondPixels]
https://github.com/JunaidCS032/MOTBeyondPixels

Submitted on 19 Sep. 2017 19:43 by
Krishna Murthy Jatavallabhula (International Institute of Information Technology, Hyderabad, India)

Running time:0.3 s
Environment:1 core @ 2.5 Ghz (C/C++)

Method Description:
This paper introduces geometry and object
shape
and pose costs for multi-object tracking
in urban
driving scenarios. 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
framework. 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.
More qualitative and quantitative results
can be
found at the following URL:
https://junaidcs032.github.io/Geometry_Obj
ectShap
e_MOT/
Parameters:
Latex Bibtex:
@inproceedings{MOTBeyondPixels,
title={Beyond Pixels: Leveraging
Geometry and
Shape Cues for Online Multi-Object
Tracking},
author={Sharma, Sarthak and Ansari,
Junaid Ahmed
and Krishna Murthy, J. and Madhava
Krishna, K.},
booktitle = {Proceedings of the IEEE
International Conference on Robotics and
Automation (ICRA)},
year={2018}
}

Detailed Results

From all 29 test sequences, our benchmark computes the commonly used tracking metrics CLEARMOT, MT/PT/ML, identity switches, and fragmentations [1,2]. The tables below show all of these metrics.


Benchmark MOTA MOTP MODA MODP
CAR 84.24 % 85.73 % 85.60 % 88.62 %

Benchmark recall precision F1 TP FP FN FAR #objects #trajectories
CAR 88.80 % 97.95 % 93.15 % 33656 705 4247 6.34 % 38507 2382

Benchmark MT PT ML IDS FRAG
CAR 73.23 % 24.00 % 2.77 % 468 944

This table as LaTeX


[1] K. Bernardin, R. Stiefelhagen: Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. JIVP 2008.
[2] Y. Li, C. Huang, R. Nevatia: Learning to associate: HybridBoosted multi-target tracker for crowded scene. CVPR 2009.


eXTReMe Tracker