Enhanced Linear Programming Tracker

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

Short name:

ELP

Benchmark:

Description:

Linear programming based tracker that performs tracklet linking over several iterations, using motion information.

Hardware:

3.4GHz 1 Core

Detector:

Public

Processing:

Batch

Last submitted:

February 04, 2015 (2 years ago)

Published:

August 21, 2015 at 15:39:55 CET

Submissions:

2

Open source:

Yes

Project page / code:

Reference:

N. McLaughlin, J. Martinez Del Rincon, P. Miller. Enhancing Linear Programming with Motion Modeling for Multi-target Tracking. In IEEE Winter Conference on Applications of Computer Vision (WACV), 2015.

Benchmark performance:

MOTAMOTPFAFMTMLFPFNID Sw.FragSpecificationsDetector
25.071.21.37.5 % 43.8 % 7,34537,3441,3961,8043.4GHz 1 CorePublic
IDF1ID PrecisionID Recall
26.238.719.8

Detailed performance:

Sequence MOTA IDF1 MOTP FAF GT MT ML FP FN ID Sw Frag
TUD-Crossing44.929.773.20.21315.4 % 15.4 % 395145446
PETS09-S2L242.225.069.51.7427.1 % 14.3 % 7204,528323464
ETH-Jelmoli35.538.773.50.64517.8 % 31.1 % 2481,3177270
ETH-Linthescher14.717.474.90.01972.0 % 77.2 % 477,4908280
ETH-Crossing21.125.474.90.1263.8 % 65.4 % 1876498
AVG-TownCentre27.537.669.12.122610.6 % 33.6 % 9334,038210359
ADL-Rundle-114.618.572.74.13215.6 % 40.6 % 2,0335,721191187
ADL-Rundle-326.027.872.72.5446.8 % 25.0 % 1,5495,803169143
KITTI-1636.633.372.60.6170.0 % 11.8 % 1338935375
KITTI-1922.732.166.31.0626.5 % 25.8 % 1,0212,967140285
Venice-112.218.173.11.3170.0 % 41.2 % 6043,3099387

Raw data:


ELP