ALExTRAC: Affinity Learning by Exploring Temporal Reinforcement within Association Chains

AVG-TownCentre


Benchmark:

Short name:

ALExTRAC

Detector:

Public

Description:

Appearance only tracking through the use of a self-supervised affinity model.

Reference:

A. Bewley, L. Ott, F. Ramos, B. Upcroft. ALExTRAC: Affinity Learning by Exploring Temporal Reinforcement within Association Chains. In International Conference on Robotics and Automation (ICRA), (to appear) 2016.

Last submitted:

September 12, 2015 (4 years ago)

Published:

February 16, 2016 at 04:03:23 CET

Submissions:

1

Project page / code:

Open source:

No

Hardware:

2.5 GHz, 4 cores

Runtime:

3.7 Hz

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
2D MOT 201517.017.371.228 (3.9)378 (52.4)9,23339,93335.070.01.61,859 (53.1)1,872 (53.5)

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN Recall Precision FAF ID Sw. Frag
ADL-Rundle-15.113.571.3573,5035,01046.255.17.0321306
ADL-Rundle-318.116.471.83102,4205,52345.765.73.9385261
AVG-TownCentre13.321.370.031384425,63721.177.41.0118152
ETH-Crossing19.623.174.72171378321.994.40.11010
ETH-Jelmoli32.933.073.26132831,33447.481.00.68698
ETH-Linthescher15.417.874.13150947,36917.594.30.190103
KITTI-1613.912.872.1041051,27924.880.10.58073
KITTI-1913.521.066.54191,1283,26638.964.81.1227326
PETS09-S2L227.512.970.60114496,15336.288.61.0385359
TUD-Crossing51.241.073.0223945958.394.30.24050
Venice-112.512.271.9077573,12031.665.61.7117134

Raw data: