Short name:
LP2D_16
Benchmark:
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.
Hardware:
3 GHz, 1 core
Detector:
Public
Processing:
Batch
Last submitted:
February 28, 2017 (1 year ago)
Published:
February 28, 2017 at 12:16:56 CET
Submissions:
1
Open source:
Yes
Project page / code:
Reference:
MOT baseline: Linear programming on 2D image coordinates.
Benchmark performance:
MOTA | MOTP | FAF | MT | ML | FP | FN | ID Sw. | Frag | Specifications | Detector |
35.7 | 75.8 | 0.9 | 8.7 % | 50.7 % | 5,084 | 111,163 | 915 | 1,264 | 3 GHz, 1 core | Public |
IDF1 | ID Precision | ID Recall |
34.2 | 58.0 | 24.2 |
Detailed performance:
Sequence | MOTA | IDF1 | MOTP | FAF | GT | MT | ML | FP | FN | ID Sw | Frag |
MOT16-01 | 23.6 | 30.3 | 73.2 | 0.2 | 23 | 13.0 % | 56.5 % | 88 | 4,783 | 17 | 39 |
MOT16-03 | 41.3 | 34.8 | 75.9 | 1.8 | 148 | 10.8 % | 31.8 % | 2,719 | 58,275 | 418 | 546 |
MOT16-06 | 41.4 | 46.4 | 73.9 | 0.4 | 221 | 10.9 % | 48.9 % | 431 | 6,202 | 126 | 165 |
MOT16-07 | 31.0 | 32.2 | 74.2 | 1.1 | 54 | 7.4 % | 48.1 % | 558 | 10,619 | 91 | 184 |
MOT16-08 | 26.5 | 29.3 | 79.8 | 0.9 | 63 | 9.5 % | 46.0 % | 575 | 11,641 | 80 | 89 |
MOT16-12 | 37.9 | 49.5 | 77.3 | 0.3 | 86 | 10.5 % | 54.7 % | 299 | 4,832 | 18 | 29 |
MOT16-14 | 16.7 | 20.9 | 74.9 | 0.6 | 164 | 2.4 % | 70.1 % | 414 | 14,811 | 165 | 212 |
Raw data: