Benchmark:
DAMUNT |
Short name:
DA_MOT
Detector:
Public
Description:
Reference:
C. Abeysinghe, C. Reid, H. Rezatofighi, B. Meyer. Tracking Different Ant Species: An Unsupervised Domain Adaptation Framework and a Dataset for Multi-object Tracking. In Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, IJCAI-23, 2023.
Last submitted:
July 05, 2023 (1 year ago)
Published:
July 06, 2023 at 06:35:30 CET
Submissions:
1
Project page / code:
Open source:
Yes
Hardware:
28 CPUs (2.2GHZ), Tesla T4
Runtime:
4.5 Hz
Benchmark performance:
Sequence | MOTA | IDF1 | HOTA | MT | ML | FP | FN | Rcll | Prcn | AssA | DetA | AssRe | AssPr | DetRe | DetPr | LocA | FAF | ID Sw. | Frag |
DAMUNT | 56.1 | 52.1 | 48.2 | 87 (42.9) | 14 (6.9) | 9,796 | 15,952 | 73.7 | 82.0 | 49.8 | 48.8 | 55.0 | 73.5 | 59.0 | 65.6 | 81.7 | 2.4 | 851 (0.0) | 1,120 (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 |
CU10L1B6In_0 | -14.2 | 19.5 | 16.2 | 0 | 5 | 1,564 | 1,450 | 48.6 | 46.7 | 10.8 | 24.5 | 11.0 | 77.9 | 38.6 | 37.1 | 79.0 | 2.2 | 210 | 219 |
CU15L1B1In_0 | 92.3 | 89.2 | 79.4 | 32 | 2 | 234 | 514 | 94.7 | 97.5 | 78.4 | 80.6 | 83.8 | 87.3 | 85.1 | 87.6 | 88.5 | 0.3 | 5 | 24 |
CU15L1B4In_0 | -1.6 | 50.0 | 42.6 | 29 | 2 | 4,777 | 899 | 84.0 | 49.7 | 48.8 | 37.3 | 60.9 | 65.5 | 70.0 | 41.4 | 82.6 | 6.6 | 26 | 39 |
CU20L1B1Out_0 | 76.2 | 73.7 | 64.5 | 15 | 1 | 1,350 | 1,830 | 86.4 | 89.6 | 62.5 | 66.6 | 67.1 | 77.0 | 75.3 | 78.1 | 86.8 | 1.9 | 17 | 19 |
CU20L1B4Out_0 | 68.0 | 37.9 | 32.3 | 9 | 0 | 1,164 | 3,280 | 77.2 | 90.5 | 21.2 | 49.4 | 23.4 | 53.7 | 54.9 | 64.3 | 74.9 | 1.6 | 168 | 210 |
CU30L1B6Out_0 | 37.6 | 20.9 | 14.7 | 2 | 4 | 707 | 7,979 | 45.3 | 90.3 | 7.8 | 28.4 | 8.5 | 44.6 | 30.2 | 60.2 | 71.5 | 1.4 | 425 | 609 |
Raw data: