Benchmark:
MOT15 |
Short name:
EDR15
Detector:
Public
Description:
The results were obtained based on this reference work together with additional appereance features (person ReID).
Reference:
Z. Fu, X. Lai, S. Naqvi. Enhanced Detection Reliability for Human Tracking Based Video Analytics. In 2019 22th International Conference on Information Fusion (FUSION), 2019.
Last submitted:
September 03, 2019 (5 years ago)
Published:
November 19, 2020 at 12:58:26 CET
Submissions:
1
Project page / code:
n/a
Open source:
No
Hardware:
GTX 1060
Runtime:
12.9 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 | 30.8 | 41.6 | 32.3 | 80 (11.1) | 253 (35.1) | 11,531 | 30,264 | 50.7 | 73.0 | 31.2 | 33.8 | 36.2 | 63.0 | 40.4 | 58.1 | 77.8 | 2.0 | 743 (14.6) | 2,283 (45.0) |
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 | 11.3 | 38.5 | 30.1 | 9 | 1 | 4,361 | 3,804 | 59.1 | 55.8 | 28.7 | 31.8 | 32.9 | 62.1 | 46.7 | 44.1 | 76.6 | 8.7 | 93 | 292 |
| ADL-Rundle-3 | 25.7 | 37.9 | 33.1 | 4 | 8 | 2,481 | 4,972 | 51.1 | 67.7 | 30.9 | 35.9 | 34.3 | 70.3 | 44.7 | 59.2 | 82.3 | 4.0 | 105 | 232 |
| AVG-TownCentre | 31.5 | 51.6 | 35.7 | 30 | 65 | 1,182 | 3,582 | 49.9 | 75.1 | 39.3 | 32.8 | 45.5 | 62.4 | 37.8 | 56.9 | 73.7 | 2.6 | 130 | 395 |
| ETH-Crossing | 23.7 | 32.9 | 25.3 | 2 | 12 | 67 | 693 | 30.9 | 82.2 | 26.4 | 24.3 | 34.4 | 58.1 | 26.1 | 69.4 | 81.2 | 0.3 | 5 | 26 |
| ETH-Jelmoli | 47.9 | 49.0 | 37.7 | 9 | 11 | 335 | 961 | 62.1 | 82.5 | 33.0 | 43.4 | 47.4 | 53.3 | 50.0 | 66.4 | 80.0 | 0.8 | 26 | 81 |
| ETH-Linthescher | 25.0 | 31.5 | 25.7 | 6 | 131 | 178 | 6,475 | 27.5 | 93.2 | 31.0 | 21.3 | 37.2 | 60.5 | 22.0 | 74.7 | 79.9 | 0.1 | 47 | 243 |
| KITTI-16 | 41.3 | 59.7 | 39.6 | 2 | 2 | 219 | 761 | 55.3 | 81.1 | 41.6 | 37.7 | 44.8 | 66.3 | 42.6 | 62.5 | 75.7 | 1.0 | 18 | 75 |
| KITTI-19 | 32.7 | 51.6 | 35.2 | 6 | 14 | 1,104 | 2,444 | 54.3 | 72.4 | 35.6 | 35.1 | 40.8 | 61.4 | 41.6 | 55.5 | 73.7 | 1.0 | 46 | 214 |
| PETS09-S2L2 | 52.2 | 39.5 | 29.6 | 5 | 3 | 658 | 3,731 | 61.3 | 90.0 | 20.5 | 43.0 | 24.1 | 50.6 | 47.0 | 69.0 | 77.6 | 1.5 | 217 | 532 |
| TUD-Crossing | 60.2 | 55.0 | 40.4 | 5 | 1 | 87 | 335 | 69.6 | 89.8 | 33.0 | 49.7 | 40.9 | 58.9 | 55.2 | 71.3 | 79.9 | 0.4 | 17 | 55 |
| Venice-1 | 25.4 | 37.0 | 31.1 | 2 | 5 | 859 | 2,506 | 45.1 | 70.5 | 32.9 | 29.9 | 34.9 | 71.5 | 35.8 | 56.0 | 78.0 | 1.9 | 39 | 138 |
Raw data: