End-to-end Data Association with Graph Neural Network for Online Multiple-Object Tracking
Short name:
EDA_GNN
Benchmark:
Description:
n/a
Hardware:
NVIDIA TITAN XP
Detector:
Public
Processing:
Online
Last submitted:
November 15, 2018 (5 months ago)
Published:
November 15, 2018 at 17:07:05 CET
Submissions:
1
Open source:
No
Project page / code:
n/a
Reference:
Paper ID 2713
Benchmark performance:
| MOTA | MOTP | FAF | MT | ML | FP | FN | ID Sw. | Frag | Specifications | Detector |
| 21.8 | 70.5 | 2.1 | 9.0 % | 40.2 % | 11,970 | 34,587 | 1,488 | 1,851 | NVIDIA TITAN XP | Public |
| IDF1 | ID Precision | ID Recall |
| 27.8 | 35.8 | 22.6 |
Detailed performance:
| Sequence | MOTA | IDF1 | MOTP | FAF | GT | MT | ML | FP | FN | ID Sw | Frag |
| TUD-Crossing | 59.2 | 36.8 | 72.4 | 0.5 | 13 | 38.5 % | 15.4 % | 105 | 298 | 47 | 35 |
| PETS09-S2L2 | 36.5 | 23.8 | 69.1 | 2.9 | 42 | 4.8 % | 4.8 % | 1,254 | 4,577 | 289 | 406 |
| ETH-Jelmoli | 35.6 | 42.8 | 72.4 | 1.0 | 45 | 24.4 % | 31.1 % | 423 | 1,144 | 66 | 62 |
| ETH-Linthescher | 20.3 | 25.8 | 73.1 | 0.3 | 197 | 4.1 % | 69.5 % | 330 | 6,700 | 87 | 96 |
| ETH-Crossing | 20.4 | 31.7 | 73.7 | 0.4 | 26 | 11.5 % | 53.8 % | 94 | 693 | 11 | 15 |
| AVG-TownCentre | 13.8 | 32.7 | 68.6 | 4.0 | 226 | 9.7 % | 32.7 % | 1,806 | 4,008 | 347 | 419 |
| ADL-Rundle-1 | 11.9 | 29.2 | 71.9 | 5.3 | 32 | 18.8 % | 34.4 % | 2,646 | 5,442 | 115 | 156 |
| ADL-Rundle-3 | 24.7 | 25.7 | 71.6 | 3.6 | 44 | 6.8 % | 20.5 % | 2,243 | 5,246 | 166 | 133 |
| KITTI-16 | 28.5 | 26.6 | 71.4 | 1.4 | 17 | 0.0 % | 11.8 % | 285 | 819 | 113 | 117 |
| KITTI-19 | 11.1 | 27.4 | 66.3 | 1.5 | 62 | 4.8 % | 30.6 % | 1,628 | 2,948 | 175 | 293 |
| Venice-1 | 13.7 | 21.8 | 70.8 | 2.6 | 17 | 11.8 % | 35.3 % | 1,156 | 2,712 | 72 | 119 |
Raw data:
n/a
