Continuous Energy Minimization for Multi-target Tracking

TUD-Crossing PETS09-S2L2 ETH-Jelmoli ETH-Linthescher ETH-Crossing AVG-TownCentre ADL-Rundle-1 ADL-Rundle-3 KITTI-16 KITTI-19

Short name:

CEM

Benchmark:

Description:

Contrary to recent approaches, we focus on designing a continuous energy that corresponds
to a more complete representation of the problem, rather than one that
is amenable to global optimization. Besides the image evidence, the
energy function takes into account physical constraints, such as target
dynamics, mutual exclusion, and track persistence. In addition, partial
image evidence is handled with explicit occlusion reasoning, and
different targets are disambiguated with an appearance model. To
nevertheless find strong local minima of the proposed non-convex energy
we construct a suitable optimization scheme that alternates between
continuous conjugate gradient descent and discrete trans-dimensional
jump moves. These moves, which are executed such that they always reduce
the energy, allow the search to escape weak minima and explore a much
larger portion of the search space of varying dimensionality.

[Parameters]
wtEdet=0.56906
wtEdyn=2.1107
wtEexc=0.86348
wtEper=1.4427
wtEreg=0.5
wtEapp=0
lambda=0.19852

Hardware:

2.6 GHz, 16 Cores

Detector:

Public

Processing:

Batch

Last submitted:

February 24, 2015 (2 years ago)

Published:

November 01, 2014 at 03:09:33 CET

Submissions:

1

Open source:

Yes

Project page / code:

Reference:

A. Milan, S. Roth, K. Schindler. Continuous Energy Minimization for Multitarget Tracking. In IEEE TPAMI, 2014.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
19.370.72.58.5 % 46.5 % 14,18034,5918131,0232.6 GHz, 16 CoresPublic
IDF1ID PrecisionID Recall
0.00.00.0

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing61.60.073.20.21330.8 % 15.4 % 483472821
PETS09-S2L244.90.070.21.54211.9 % 14.3 % 6574,506150165
ETH-Jelmoli36.20.071.71.14513.3 % 28.9 % 4801,1102854
ETH-Linthescher18.40.073.80.31975.1 % 72.1 % 3286,8838077
ETH-Crossing18.20.071.40.42611.5 % 57.7 % 79733810
AVG-TownCentre-2.60.068.94.82265.8 % 54.0 % 2,1704,979186232
ADL-Rundle-19.20.071.47.83221.9 % 12.5 % 3,9044,46582125
ADL-Rundle-315.40.071.75.24413.6 % 20.5 % 3,2465,2827683
KITTI-1631.60.072.31.3175.9 % 17.6 % 2788592632
KITTI-199.90.065.61.7624.8 % 24.2 % 1,8532,86792175
Venice-117.70.071.02.51717.6 % 23.5 % 1,1372,5605749

Raw data:


CEM