ApLift: Approximate Lifted Disjoint Paths Tracker
AVG-TownCentre
Short name:
ApLift
Detector:
Public
Description:
Reference:
A. Hornakova*, T. Kaiser*, M. Rolinek, B. Rosenhahn, P. Swoboda, R. Henschel. Making Higher Order MOT Scalable: An Efficient Approximate Solver for Lifted Disjoint Paths. In International Conference on Computer Vision (ICCV), 2021.
Last submitted:
March 14, 2021 (3 years ago)
Published:
March 22, 2021 at 11:46:43 CET
Submissions:
1
Project page / code:
Open source:
No
Hardware:
5GHZ, 5 cores
Runtime:
0.6 Hz
Benchmark performance:
| Sequence | MOTA | IDF1 | HOTA | MT | ML | FP | FN | Rcll | Prcn | AssA | DetA | AssRe | AssPr | DetRe | DetPr | LocA | FAF | ID Sw. | Frag |
| MOT15 | 51.1 | 59.0 | 45.7 | 284 (39.4) | 163 (22.6) | 10,070 | 19,288 | 68.6 | 80.7 | 46.7 | 45.4 | 55.2 | 68.7 | 54.1 | 63.7 | 79.3 | 1.7 | 677 (0.0) | 1,022 (0.0) |
Detailed performance:
| Sequence | MOTA | IDF1 | HOTA | MT | ML | FP | FN | Rcll | Prcn | AssA | DetA | AssRe | AssPr | DetRe | DetPr | LocA | FAF | ID Sw. | Frag |
| ADL-Rundle-1 | 27.8 | 55.9 | 42.9 | 17 | 1 | 4,031 | 2,636 | 71.7 | 62.3 | 47.7 | 38.8 | 53.9 | 68.3 | 56.1 | 48.7 | 77.3 | 8.1 | 52 | 115 |
| ADL-Rundle-3 | 57.9 | 64.5 | 52.8 | 22 | 8 | 1,008 | 3,232 | 68.2 | 87.3 | 55.7 | 50.4 | 61.7 | 77.5 | 57.1 | 73.1 | 83.8 | 1.6 | 41 | 50 |
| AVG-TownCentre | 58.5 | 63.6 | 46.0 | 93 | 29 | 563 | 2,170 | 69.6 | 89.8 | 45.5 | 47.2 | 53.3 | 64.6 | 51.9 | 67.0 | 76.0 | 1.3 | 235 | 276 |
| ETH-Crossing | 62.3 | 58.9 | 54.1 | 10 | 8 | 114 | 256 | 74.5 | 86.8 | 52.1 | 56.4 | 65.7 | 72.1 | 64.5 | 75.2 | 85.8 | 0.5 | 8 | 7 |
| ETH-Jelmoli | 55.9 | 71.1 | 53.8 | 23 | 12 | 600 | 511 | 79.9 | 77.2 | 56.8 | 51.0 | 72.0 | 64.9 | 65.2 | 63.0 | 82.4 | 1.4 | 9 | 28 |
| ETH-Linthescher | 54.1 | 59.7 | 48.4 | 68 | 86 | 1,023 | 3,038 | 66.0 | 85.2 | 49.6 | 47.4 | 66.4 | 62.8 | 54.2 | 70.0 | 82.1 | 0.9 | 41 | 81 |
| KITTI-16 | 55.1 | 66.9 | 43.7 | 3 | 1 | 177 | 566 | 66.7 | 86.5 | 44.1 | 43.4 | 49.4 | 60.1 | 48.7 | 63.1 | 74.5 | 0.8 | 21 | 52 |
| KITTI-19 | 47.8 | 61.1 | 41.9 | 18 | 13 | 824 | 1,912 | 64.2 | 80.6 | 42.8 | 41.4 | 47.5 | 65.0 | 48.0 | 60.2 | 74.7 | 0.8 | 54 | 123 |
| PETS09-S2L2 | 55.3 | 44.0 | 32.9 | 9 | 2 | 709 | 3,412 | 64.6 | 89.8 | 23.7 | 46.0 | 27.1 | 60.0 | 50.4 | 70.0 | 78.2 | 1.6 | 193 | 259 |
| TUD-Crossing | 89.3 | 84.6 | 60.4 | 13 | 0 | 65 | 41 | 96.3 | 94.2 | 55.8 | 65.5 | 64.8 | 66.1 | 72.6 | 71.0 | 78.1 | 0.3 | 12 | 11 |
| Venice-1 | 45.6 | 56.4 | 43.1 | 8 | 3 | 956 | 1,514 | 66.8 | 76.1 | 45.3 | 41.3 | 53.3 | 72.3 | 51.3 | 58.4 | 78.5 | 2.1 | 11 | 20 |
Raw data:
