LP2D: Linear Programming on 2D coordinates


Benchmark:

Short name:

LP2D

Detector:

Public

Description:

The tracker builds a graphical model where nodes are the 2D detections. The solution is then found by computing the minimum cost-flow problem.

Project page / code:

Reference:

MOT baseline: Linear programming on 2D image coordinates.

Processing:

Batch

Last submitted:

October 27, 2014 (5 years ago)

Published:

October 27, 2014 at 20:01:43 CET

Submissions:

1

Open source:

Yes

Hardware:

2.6 GHz, 16 Cores

Benchmark performance:

Sequence MOTA IDF1 MOTP MT ML FP FN ID Sw.
2D MOT 201519.80.071.248.0297.011,58036,0451,649

Detailed performance:

Sequence MOTA IDF1 MOTP MT ML FP FN ID Sw.
ADL-Rundle-12.90.072.25.07.03,8534,934252
ADL-Rundle-313.70.072.81.011.02,4615,914400
AVG-TownCentre15.50.068.519.075.01,5254,257260
ETH-Crossing21.40.076.31.017.01975910
ETH-Jelmoli40.70.073.57.012.03191,14441
ETH-Linthescher16.90.076.44.0145.01247,22277
KITTI-1635.50.072.00.02.019785347
KITTI-1920.10.065.25.013.01,3782,79397
PETS09-S2L240.70.070.24.07.08484,554319
TUD-Crossing49.50.074.12.02.04945948
Venice-111.00.072.40.06.08073,15698

Raw data:


LP2D