Benchmark:
MOT15 |
Short name:
GMPHD_15
Detector:
Public
Description:
Stage 1: low-level association between targets states and observations(i.e., detection responses), frame by frame
-We only used detection reponses with the positive confidence value among provided detection results
Stage 2: short tracklet elimination under the set length(parameter)
-Minimum tracklet length for mid-level association : 10
Stage 3: mid-level association between rough tracklets(almost fragmented or ID swtiched)
-Only Motion and size are used
-Appearance is not used
Reference:
Y. Song, M. Jeon. Online Multiple Object Tracking with the Hierarchically Adopted GM-PHD Filter using Motion and Appearance. In IEEE/IEIE The International Conference on Consumer Electronics (ICCE) Asia, 2016.
Last submitted:
December 05, 2016 (7 years ago)
Published:
December 05, 2016 at 10:21:45 CET
Submissions:
1
Project page / code:
Open source:
No
Hardware:
3.5 GHz, 4 Cores
Runtime:
19.8 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 | 18.5 | 28.4 | 21.3 | 28 (3.9) | 399 (55.3) | 7,864 | 41,766 | 32.0 | 71.4 | 21.1 | 21.7 | 25.5 | 53.7 | 24.5 | 54.7 | 74.8 | 1.4 | 459 (14.3) | 1,266 (39.5) |
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 | 10.0 | 28.4 | 22.4 | 2 | 13 | 2,366 | 5,954 | 36.0 | 58.6 | 22.8 | 22.4 | 25.2 | 61.5 | 28.1 | 45.7 | 75.6 | 4.7 | 58 | 176 |
| ADL-Rundle-3 | 15.3 | 23.8 | 16.6 | 0 | 18 | 1,374 | 7,174 | 29.4 | 68.5 | 14.0 | 19.8 | 16.2 | 60.4 | 22.6 | 52.7 | 76.4 | 2.2 | 64 | 108 |
| AVG-TownCentre | 16.3 | 28.4 | 22.4 | 9 | 147 | 413 | 5,542 | 22.5 | 79.5 | 31.6 | 16.0 | 35.7 | 64.9 | 16.9 | 60.0 | 74.5 | 0.9 | 26 | 103 |
| ETH-Crossing | 6.4 | 22.5 | 17.7 | 2 | 19 | 180 | 756 | 24.6 | 57.8 | 18.7 | 16.8 | 32.0 | 44.8 | 19.8 | 46.4 | 77.0 | 0.8 | 3 | 11 |
| ETH-Jelmoli | 34.3 | 42.8 | 30.7 | 5 | 16 | 317 | 1,335 | 47.4 | 79.1 | 29.7 | 32.1 | 40.2 | 45.8 | 36.3 | 60.7 | 76.3 | 0.7 | 16 | 59 |
| ETH-Linthescher | 16.2 | 24.0 | 20.0 | 5 | 142 | 539 | 6,896 | 22.8 | 79.1 | 24.2 | 16.6 | 32.8 | 47.9 | 17.7 | 61.3 | 75.5 | 0.5 | 44 | 89 |
| KITTI-16 | 27.5 | 37.8 | 24.7 | 0 | 5 | 154 | 1,061 | 37.6 | 80.6 | 23.5 | 25.9 | 27.4 | 50.4 | 28.5 | 61.1 | 75.9 | 0.7 | 19 | 52 |
| KITTI-19 | 11.5 | 33.3 | 24.3 | 3 | 17 | 1,285 | 3,394 | 36.5 | 60.3 | 24.8 | 24.3 | 29.6 | 50.1 | 28.7 | 47.4 | 70.4 | 1.2 | 50 | 254 |
| PETS09-S2L2 | 31.9 | 28.4 | 19.3 | 0 | 13 | 467 | 5,965 | 38.1 | 88.7 | 14.1 | 26.6 | 17.5 | 43.2 | 28.0 | 65.1 | 73.6 | 1.1 | 131 | 315 |
| TUD-Crossing | 50.5 | 49.2 | 34.4 | 2 | 2 | 41 | 485 | 56.0 | 93.8 | 31.2 | 38.1 | 36.1 | 55.6 | 40.9 | 68.5 | 76.7 | 0.2 | 19 | 29 |
| Venice-1 | 13.2 | 22.8 | 16.6 | 0 | 7 | 728 | 3,204 | 29.8 | 65.1 | 13.9 | 20.2 | 16.5 | 50.5 | 23.3 | 51.0 | 75.4 | 1.6 | 29 | 70 |
Raw data: