INARLA:
TUD-Crossing
Benchmark:
Short name:
INARLA
Detector:
Public
Description:
n/a
Reference:
H. Wu, Y. Hu, K. Wang, H. Li, L. Nie, H. Cheng. Instance-aware representation learning and association for online multi-person tracking. In Pattern Recognition, 2019.
Last submitted:
February 11, 2018 (2 years ago)
Published:
August 09, 2019 at 10:02:30 CET
Submissions:
1
Project page / code:
n/a
Open source:
No
Hardware:
CPU: 8 core, 3.7 GHz, GPU: Tesla K40
Runtime:
2.6 Hz
Benchmark performance:
| Sequence | MOTA | IDF1 | MOTP | MT | ML | FP | FN | Recall | Precision | FAF | ID Sw. | Frag |
| 2D MOT 2015 | 34.7 | 42.1 | 70.7 | 90 (12.5) | 216 (30.0) | 9,855 | 29,158 | 52.5 | 76.6 | 1.7 | 1,112 (21.2) | 2,848 (54.2) |
Detailed performance:
| Sequence | MOTA | IDF1 | MOTP | MT | ML | FP | FN | Recall | Precision | FAF | ID Sw. | Frag |
| ADL-Rundle-1 | 20.1 | 42.1 | 69.2 | 6 | 3 | 3,463 | 3,868 | 58.4 | 61.1 | 6.9 | 100 | 402 |
| ADL-Rundle-3 | 36.0 | 37.8 | 75.0 | 7 | 8 | 1,383 | 5,014 | 50.7 | 78.8 | 2.2 | 109 | 221 |
| AVG-TownCentre | 32.1 | 45.7 | 67.4 | 20 | 54 | 955 | 3,649 | 49.0 | 78.6 | 2.1 | 250 | 570 |
| ETH-Crossing | 35.2 | 48.5 | 74.9 | 2 | 11 | 72 | 566 | 43.6 | 85.9 | 0.3 | 12 | 22 |
| ETH-Jelmoli | 48.1 | 57.1 | 73.9 | 13 | 13 | 329 | 955 | 62.4 | 82.8 | 0.7 | 33 | 76 |
| ETH-Linthescher | 41.3 | 48.3 | 73.3 | 20 | 104 | 472 | 4,704 | 47.3 | 90.0 | 0.4 | 69 | 164 |
| KITTI-16 | 32.9 | 55.0 | 67.4 | 0 | 1 | 329 | 786 | 53.8 | 73.6 | 1.6 | 26 | 113 |
| KITTI-19 | 24.4 | 46.9 | 66.7 | 7 | 15 | 1,324 | 2,649 | 50.4 | 67.0 | 1.3 | 67 | 326 |
| PETS09-S2L2 | 40.5 | 26.1 | 69.9 | 3 | 5 | 703 | 4,657 | 51.7 | 87.6 | 1.6 | 380 | 743 |
| TUD-Crossing | 73.0 | 63.3 | 73.3 | 7 | 0 | 12 | 260 | 76.4 | 98.6 | 0.1 | 25 | 51 |
| Venice-1 | 36.4 | 40.1 | 69.0 | 5 | 2 | 813 | 2,050 | 55.1 | 75.6 | 1.8 | 41 | 160 |
Raw data:
