CRFTrack_:
TUD-Crossing
Benchmark:
Short name:
CRFTrack_
Detector:
Public
Description:
n/a
Reference:
Jun xiang, Chao Ma, Guohan Xu, Jianhua Hou, End-to-End Learning Deep CRF models for Multi-Object Tracking. In IEEE Transactions on Circuits and Systems for Video Technology, 2020
Last submitted:
June 25, 2019 (1 year ago)
Published:
June 26, 2019 at 11:13:54 CET
Submissions:
1
Project page / code:
n/a
Open source:
Yes
Hardware:
2.5GHZ
Runtime:
3.2 Hz
Benchmark performance:
| Sequence | MOTA | IDF1 | MOTP | MT | ML | FP | FN | Recall | Precision | FAF | ID Sw. | Frag |
| 2D MOT 2015 | 40.0 | 49.6 | 71.9 | 166 (23.0) | 206 (28.6) | 10,295 | 25,917 | 57.8 | 77.5 | 1.8 | 658 (11.4) | 1,508 (26.1) |
Detailed performance:
| Sequence | MOTA | IDF1 | MOTP | MT | ML | FP | FN | Recall | Precision | FAF | ID Sw. | Frag |
| ADL-Rundle-1 | 21.8 | 43.4 | 70.7 | 11 | 2 | 3,756 | 3,445 | 63.0 | 60.9 | 7.5 | 75 | 176 |
| ADL-Rundle-3 | 38.2 | 45.1 | 75.8 | 6 | 8 | 1,568 | 4,623 | 54.5 | 77.9 | 2.5 | 92 | 124 |
| AVG-TownCentre | 49.0 | 63.3 | 68.7 | 66 | 43 | 778 | 2,789 | 61.0 | 84.9 | 1.7 | 75 | 353 |
| ETH-Crossing | 31.9 | 52.1 | 76.5 | 8 | 11 | 126 | 553 | 44.9 | 78.1 | 0.6 | 4 | 9 |
| ETH-Jelmoli | 48.5 | 66.4 | 76.2 | 16 | 13 | 440 | 857 | 66.2 | 79.2 | 1.0 | 9 | 46 |
| ETH-Linthescher | 34.6 | 46.6 | 78.2 | 21 | 114 | 252 | 5,561 | 37.7 | 93.0 | 0.2 | 28 | 71 |
| KITTI-16 | 51.6 | 66.1 | 71.7 | 4 | 1 | 184 | 617 | 63.7 | 85.5 | 0.9 | 22 | 56 |
| KITTI-19 | 39.8 | 58.1 | 67.9 | 9 | 10 | 900 | 2,267 | 57.6 | 77.4 | 0.9 | 52 | 201 |
| PETS09-S2L2 | 51.0 | 30.9 | 69.4 | 9 | 3 | 1,073 | 3,370 | 65.0 | 85.4 | 2.5 | 282 | 393 |
| TUD-Crossing | 80.8 | 88.8 | 70.2 | 10 | 0 | 42 | 163 | 85.2 | 95.7 | 0.2 | 7 | 25 |
| Venice-1 | 37.3 | 60.1 | 71.4 | 6 | 1 | 1,176 | 1,672 | 63.4 | 71.1 | 2.6 | 12 | 54 |
Raw data:
