Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism

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

Short name:

AM

Benchmark:

Description:

n/a

Hardware:

2.4 GHZ, 1 Core, TITAN X

Detector:

Public

Processing:

Online

Last submitted:

March 17, 2017 (7 months ago)

Published:

August 13, 2017 at 15:24:27 CET

Submissions:

2

Open source:

No

Project page / code:

n/a

Reference:

Q. Chu, W. Ouyang, H. Li, X. Wang, B. Liu, N. Yu. Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism. In arXiv preprint arXiv:1708.02843, 2017.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
34.370.50.911.4 % 43.4 % 5,15434,8483481,4632.4 GHZ, 1 Core, TITAN XPublic
IDF1ID PrecisionID Recall
48.370.836.6

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing68.681.774.40.11346.2 % 15.4 % 25315622
PETS09-S2L247.744.369.21.64216.7 % 14.3 % 7184,206115356
ETH-Jelmoli43.159.873.40.64520.0 % 37.8 % 2771,1551143
ETH-Linthescher21.132.473.60.11973.6 % 75.6 % 1636,8721554
ETH-Crossing26.542.374.10.1263.8 % 57.7 % 16721012
AVG-TownCentre37.553.968.11.422614.2 % 30.5 % 6453,74279332
ADL-Rundle-127.845.471.42.03218.8 % 46.9 % 1,0005,69726181
ADL-Rundle-339.352.472.01.24411.4 % 29.5 % 7205,41836109
KITTI-1640.660.670.60.9175.9 % 17.6 % 1868051957
KITTI-1924.350.166.31.0628.1 % 24.2 % 1,0982,90937239
Venice-127.347.871.30.71717.6 % 52.9 % 3063,008458

Raw data:


AM