PathTrack: Fast Trajectory Annotation with Path Supervision

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

Short name:

TRID

Benchmark:

Description:

n/a

Hardware:

1 CPU + 1 GPU

Detector:

Private

Processing:

Batch

Last submitted:

September 18, 2016 (3 years ago)

Published:

April 19, 2017 at 14:46:56 CET

Submissions:

1

Open source:

No

Project page / code:

Reference:

S. Manen, M. Gygli, D. Dai, L. Van Gool. PathTrack: Fast Trajectory Annotation with Path Supervision. In ArXiv e-prints, 2017.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
55.776.51.140.6 % 25.8 % 6,27320,6113516671 CPU + 1 GPUPrivate
IDF1ID PrecisionID Recall
61.070.353.9

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing80.881.073.30.11369.2 % 7.7 % 25179811
PETS09-S2L265.647.774.51.24228.6 % 7.1 % 5022,692121185
ETH-Jelmoli51.769.278.71.24546.7 % 37.8 % 528696216
ETH-Linthescher65.067.080.00.719742.1 % 35.0 % 7842,3133055
ETH-Crossing63.273.885.10.22630.8 % 50.0 % 3932558
AVG-TownCentre57.970.772.41.522642.9 % 22.1 % 6622,26980156
ADL-Rundle-147.259.475.92.53237.5 % 21.9 % 1,2433,6511833
ADL-Rundle-357.458.182.20.84436.4 % 15.9 % 4813,8143935
KITTI-1645.568.269.51.81741.2 % 5.9 % 3815301635
KITTI-1942.164.370.31.36237.1 % 16.1 % 1,3421,72428128
Venice-140.750.277.20.61729.4 % 47.1 % 2862,41845

Raw data:


TRID