MOT20 Results

Download all results for this benchmark

Click on a measure to sort the table accordingly. See below for a more detailed description.


Showing only entries that use public detections!

Benchmark Statistics

TrackerMOTAIDF1HOTAMTMLFPFNRcllPrcnAssADetAAssReAssPrDetReDetPrLocAFAFID Sw.FragHz
SCOSORT
1. online method using public detections 		76.9 76.9 62.8 846 (68.1)126 (10.1)20,875 96,944 81.3 95.3 62.5 63.4 67.7 79.2 68.3 80.1 84.3 4.7 1,524 (0.0)2,986 (0.0)17.5
FeatureSORT
2. online method using public detections 		76.6 75.1 61.3 832 (67.0)142 (11.4)25,083 95,027 81.6 94.4 60.1 62.7 66.3 76.1 68.1 78.7 83.6 5.6 1,081 (0.0)1,250 (0.0)4.8
H. Hashempoor, R. Koikara, Y. Hwang. FeatureSORT: Essential Features for Effective Tracking. In , 2024.
kalman_pub
3. online method using public detections 		67.0 70.2 56.4 592 (47.7)263 (21.2)9,685 160,303 69.0 97.4 58.3 54.8 65.0 73.6 57.7 81.4 84.1 2.2 680 (0.0)1,738 (0.0)17.7
Y. Zhang, P. Sun, Y. Jiang, D. Yu, F. Weng, Z. Yuan, P. Luo, W. Liu, X. Wang. ByteTrack: Multi-Object Tracking by Associating Every Detection Box. In Proceedings of the European Conference on Computer Vision (ECCV), 2022.
OUTrack_fm_p
4. online method using public detections 		65.4 65.1 52.1 615 (49.5)165 (13.3)38,243 137,770 73.4 90.8 50.7 53.8 53.9 77.8 59.4 73.6 81.4 8.5 2,885 (0.0)7,205 (0.0)5.1
Q. Liu, D. Chen, Q. Chu, L. Yuan, B. Liu, L. Zhang, N. Yu. Online Multi-Object Tracking with Unsupervised Re-Identification Learning and Occlusion Estimation. In Neurocomputing, 2022.
UTM
5. online method using public detections 		64.4 65.9 53.3 807 (65.0)108 (8.7)82,726 98,974 80.9 83.5 51.0 56.1 60.6 65.7 66.9 69.0 82.3 18.5 2,592 (0.0)1,735 (0.0)6.2
S. You, H. Yao, k. Bao, C. Xu. UTM: A Unified Multiple Object Tracking Model with Identity-Aware Feature Enhancement. In CVPR, 2023.
RETracker
6. online method using public detections 		62.4 53.0 45.3 605 (48.7)192 (15.5)43,503 147,451 71.5 89.5 38.6 53.4 53.3 52.3 59.3 74.2 82.1 9.7 3,804 (0.0)9,245 (0.0)0.0
Y. Kawanishi. Label-Based Multiple Object Ensemble Tracking with Randomized Frame Dropping. In Proceedings of the 26th International Conference on Pattern Recognition, 2022.
SUSHI
7. using public detections 		61.6 71.6 55.4 591 (47.6)239 (19.2)29,429 168,098 67.5 92.2 60.6 50.8 65.8 76.7 55.2 75.4 81.8 6.6 1,053 (0.0)1,471 (0.0)5.3
O. Cetintas, G. Brasó, L. Leal-Taixé. Unifying Short and Long-Term Tracking with Graph Hierarchies. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2023.
TransCtr
8. online method using public detections 		61.0 49.8 43.5 601 (48.4)192 (15.5)49,189 147,890 71.4 88.3 36.1 52.9 44.5 59.0 59.2 73.1 81.7 11.0 4,493 (0.0)8,950 (0.0)1.0
Y. Xu, Y. Ban, G. Delorme, C. Gan, D. Rus, X. Alameda-Pineda. TransCenter: Transformers with Dense Queries for Multiple-Object Tracking. In arXiv, 2021.
MPTC
9. online method using public detections 		60.6 59.7 48.5 635 (51.1)208 (16.7)45,318 153,978 70.2 88.9 46.5 50.7 51.6 66.4 56.8 72.0 81.4 10.1 4,533 (64.5)5,163 (73.5)0.7
D. Stadler, J. Beyerer. Multi-Pedestrian Tracking with Clusters. In 2021 17th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS), 2021.
TMOH
10. online method using public detections 		60.1 61.2 48.9 580 (46.7)221 (17.8)38,043 165,899 67.9 90.2 48.4 49.7 52.9 71.0 54.8 72.9 81.2 8.5 2,342 (34.5)4,320 (63.6)0.6
D. Stadler, J. Beyerer. Improving Multiple Pedestrian Tracking by Track Management and Occlusion Handling. In CVPR, 2021.
TrackerMOTAIDF1HOTAMTMLFPFNRcllPrcnAssADetAAssReAssPrDetReDetPrLocAFAFID Sw.FragHz
OCSORTpublic
11. online method using public detections 		59.9 67.0 54.3 478 (38.5)330 (26.6)4,434 202,502 60.9 98.6 59.5 49.7 65.1 76.6 51.6 83.5 85.0 1.0 554 (0.0)2,345 (0.0)27.6
J. Cao, X. Weng, R. Khirodkar, J. Pang, K. Kitani. Observation-Centric SORT: Rethinking SORT for Robust Multi-Object Tracking. In , 2022.
mfi_tst
12. online method using public detections 		59.3 59.1 47.1 511 (41.1)215 (17.3)36,150 172,782 66.6 90.5 46.2 48.4 50.0 73.9 53.1 72.2 79.9 8.1 1,919 (0.0)3,743 (0.0)0.5
J. Y, H. Ge, J. Yang, Y. Tong, S. Su. Online multi-object tracking using multi-function integration and tracking simulation training. In Applied Intelligence, 2021.
MOTer
13. online method using public detections 		59.1 49.9 43.6 598 (48.1)189 (15.2)59,624 147,333 71.5 86.1 37.0 51.8 45.2 58.8 58.9 71.0 81.2 13.3 4,597 (0.0)9,232 (0.0)1.0
Y. Xu, Y. Ban, G. Delorme, C. Gan, D. Rus, X. Alameda-Pineda. TransCenter: Transformers with Dense Queries for Multiple-Object Tracking. In arXiv, 2021.
ApLift
14. using public detections 		58.9 56.5 46.6 513 (41.3)264 (21.3)17,739 192,736 62.8 94.8 45.2 48.2 48.1 76.8 51.3 77.5 82.2 4.0 2,241 (0.0)2,112 (0.0)0.4
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.
MPNTrack
15. using public detections 		57.6 59.1 46.8 474 (38.2)279 (22.5)16,953 201,384 61.1 94.9 47.3 46.6 52.7 70.0 49.5 76.9 81.6 3.8 1,210 (19.8)1,420 (23.2)6.5
G. Braso, L. Leal-Taixe. Learning a Neural Solver for Multiple Object Tracking. In CVPR, 2020.
IQHAT
16. using public detections 		57.1 57.7 45.7 507 (40.8)249 (20.0)32,247 187,937 63.7 91.1 44.7 46.9 50.1 68.8 51.1 73.1 80.5 7.2 1,875 (0.0)2,543 (0.0)7.5
Y. He, X. Wei, X. Hong, W. Ke, Y. Gong. Identity-Quantity Harmonic Multi-Object Tracking. In IEEE Transactions on Image Processing, 2022.
LPC_MOT
17. using public detections 		56.3 62.5 49.0 424 (34.1)313 (25.2)11,726 213,056 58.8 96.3 52.4 45.8 54.7 81.3 48.1 78.7 82.3 2.6 1,562 (26.6)1,865 (31.7)0.7
P. Dai, R. Weng, W. Choi, C. Zhang, Z. He, W. Ding. Learning a Proposal Classifier for Multiple Object tracking. In CVPR (Accepted), 2021.
Sp_Con
18. using public detections 		54.6 53.4 42.5 407 (32.8)317 (25.5)9,486 223,607 56.8 96.9 41.4 43.8 48.2 63.8 46.0 78.4 82.0 2.1 1,674 (0.0)2,455 (0.0)1.4
G. Wang, Y. Wang, R. Gu, W. Hu, J. Hwang. Split and Connect: A Universal Tracklet Booster for Multi-Object Tracking. In IEEE Transactions on Multimedia, 2022.
GNNMatch
19. online method using public detections 		54.5 49.0 40.2 407 (32.8)317 (25.5)9,522 223,611 56.8 96.9 37.0 43.8 44.0 58.7 45.9 78.4 82.0 2.1 2,038 (35.9)2,456 (43.3)0.1
I. Papakis, A. Sarkar, A. Karpatne. GCNNMatch: Graph Convolutional Neural Networks for Multi-Object Tracking via Sinkhorn Normalization. In arXiv preprint arXiv:2010.00067, 2020.
MOT20_TBC
20. using public detections 		54.5 50.1 0.0 415 (33.4)245 (19.7)37,937 195,242 62.3 89.5 0.0 0.0 0.0 0.0 0.0 0.0 100.0 8.5 2,449 (39.3)2,580 (41.4)5.6
Tracking-by-Counting: Using Network Flows on Crowd Density Maps for Tracking Multiple Targets[J]. IEEE Transactions on Image Processing, 2020, 30: 1439-1452.
TrackerMOTAIDF1HOTAMTMLFPFNRcllPrcnAssADetAAssReAssPrDetReDetPrLocAFAFID Sw.FragHz
UnsupTrack
21. online method using public detections 		53.6 50.6 41.7 376 (30.3)311 (25.0)6,439 231,298 55.3 97.8 40.2 43.3 43.2 75.9 45.1 79.8 82.6 1.4 2,178 (39.4)4,335 (78.4)1.3
S. Karthik, A. Prabhu, V. Gandhi. Simple Unsupervised Multi-Object Tracking. In Arxiv, 2020.
Tracktor++v2
22. online method using public detections 		52.6 52.7 42.1 365 (29.4)331 (26.7)6,930 236,680 54.3 97.6 42.0 42.3 45.9 71.6 44.1 79.3 82.4 1.5 1,648 (30.4)4,374 (80.6)1.2
P. Bergmann, T. Meinhardt, L. Leal-Taixé. Tracking without bells and whistles. In ICCV, 2019.
SFS
23. online method using public detections 		50.8 41.1 32.7 341 (27.5)251 (20.2)50,139 200,932 61.2 86.3 25.1 43.1 34.5 47.4 48.0 67.8 78.2 11.2 3,503 (57.3)7,617 (124.5)0.1
MTSFS： Online Multi-Object Tracking Based on Salient Feature Selection in Crowded Scenes
BBT
24. using public detections 		46.8 42.2 35.8 312 (25.1)289 (23.3)35,014 236,176 54.4 88.9 32.2 40.2 35.1 70.2 43.8 71.7 80.6 7.8 3,880 (71.4)7,207 (132.6)8.0
MTAP-D-20-01870 Tracking Subjects and Detecting Relationships in Crowded City Videos (under review)
FlowTracker
25. online method using public detections 		46.7 42.4 35.6 345 (27.8)249 (20.0)54,732 217,371 58.0 84.6 31.2 41.3 34.2 68.3 46.4 67.6 79.1 12.2 3,532 (0.0)5,165 (0.0)19.2
H. Nishimura et al., "SDOF-Tracker: Fast and Accurate Multiple Human Tracking by Skipped-Detection and Optical-Flow"
IOU_KMM
26. online method using public detections 		46.5 49.4 40.4 371 (29.9)244 (19.6)57,517 214,777 58.5 84.0 39.4 41.9 41.7 75.2 47.3 67.9 79.9 12.8 4,509 (0.0)7,557 (0.0)30.3
O. Urbann, O. Bredtmann, M. Otten, P. Richter, T. Bauer, D. Zibriczky. Online and Real-Time Tracking in a Surveillance Scenario. In arXiv, 2021.
CT_v0
27. using public detections 		45.1 35.6 33.0 409 (32.9)235 (18.9)69,491 207,927 59.8 81.7 26.3 42.3 27.7 75.4 48.7 66.4 79.8 15.5 6,492 (0.0)6,351 (0.0)317.9
J. Lohn-Jaramillo, L. Ray, R. Granger, E. Bowen. ClusterTracker: An Efficiency-Focused Multiple Object Tracking Method. In , 2022.
GMPHD_Rd20
28. online method using public detections 		44.7 43.5 35.6 293 (23.6)274 (22.1)42,778 236,116 54.4 86.8 32.0 39.9 34.1 72.0 44.0 70.3 80.2 9.6 7,492 (137.8)11,153 (205.1)25.2
N. Baisa. Occlusion-robust online multi-object visual tracking using a GM-PHD filter with CNN-based re-identification. In Journal of Visual Communication and Image Representation, 2021.
SORT20
29. online method using public detections 		42.7 45.1 36.1 208 (16.7)326 (26.2)27,521 264,694 48.8 90.2 35.9 36.7 39.4 66.9 39.5 73.0 81.0 6.1 4,470 (91.5)17,798 (364.4)57.3
A. Bewley, Z. Ge, L. Ott, F. Ramos, B. Upcroft. Simple online and realtime tracking. In 2016 IEEE International Conference on Image Processing (ICIP), 2016.
OVBT17
30. online method using public detections 		40.0 37.8 30.5 141 (11.4)374 (30.1)23,368 282,949 45.3 90.9 27.4 34.4 31.0 60.6 36.7 73.7 81.1 5.2 4,210 (0.0)10,026 (0.0)0.1
Y. Ban, S. Ba, X. Alameda-Pineda, R. Horaud. Tracking multiple persons based on a variational bayesian model. In European Conference on Computer Vision Workshops, 2016.
SequencesFramesTrajectoriesBoxes
444791501765465

Evaluation Measures

Lower is better. Higher is better.
Measure Better Perfect Description
MOTA higher 100%Multi-Object Tracking Accuracy (+/- denotes standard deviation across all sequences) [1]. This measure combines three error sources: false positives, missed targets and identity switches.
IDF1 higher 100%ID F1 Score [2]. The ratio of correctly identified detections over the average number of ground-truth and computed detections.
HOTA higher 100%Higher Order Tracking Accuracy [3]. Geometric mean of detection accuracy and association accuracy. Averaged across localization thresholds.
MT higher 100%Mostly tracked targets. The ratio of ground-truth trajectories that are covered by a track hypothesis for at least 80% of their respective life span.
ML lower 0%Mostly lost targets. The ratio of ground-truth trajectories that are covered by a track hypothesis for at most 20% of their respective life span.
FP lower 0The total number of false positives.
FN lower 0The total number of false negatives (missed targets).
Rcll higher 100%Ratio of correct detections to total number of GT boxes.
Prcn higher 100%Ratio of TP / (TP+FP).
AssA higher 100%Association Accuracy [3]. Association Jaccard index averaged over all matching detections and then averaged over localization thresholds.
DetA higher 100%Detection Accuracy [3]. Detection Jaccard index averaged over localization thresholds.
AssRe higher 100%Association Recall [3]. TPA / (TPA + FNA) averaged over all matching detections and then averaged over localization thresholds.
AssPr higher 100%Association Precision [3]. TPA / (TPA + FPA) averaged over all matching detections and then averaged over localization thresholds.
DetRe higher 100%Detection Recall [3]. TP /(TP + FN) averaged over localization thresholds.
DetPr higher 100%Detection Precision [3]. TP /(TP + FP) averaged over localization thresholds.
LocA higher 100%Localization Accuracy [3]. Average localization similarity averaged over all matching detections and averaged over localization thresholds.
FAF lower 0The average number of false alarms per frame.
ID Sw. lower 0Number of Identity Switches (ID switch ratio = #ID switches / recall) [4]. Please note that we follow the stricter definition of identity switches as described in the reference
Frag lower 0The total number of times a trajectory is fragmented (i.e. interrupted during tracking).
Hz higher Inf.Processing speed (in frames per second excluding the detector) on the benchmark. The frequency is provided by the authors and not officially evaluated by the MOTChallenge.

Legend

Symbol Description
online method This is an online (causal) method, i.e. the solution is immediately available with each incoming frame and cannot be changed at any later time.
using public detections This method used the provided detection set as input.
using private detections This method used a private detection set as input.
new This entry has been submitted or updated less than a week ago.

References:


[1] Bernardin, K. & Stiefelhagen, R. Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. Image and Video Processing, 2008(1):1-10, 2008.
[2] Ristani, E., Solera, F., Zou, R., Cucchiara, R. & Tomasi, C. Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking. In ECCV workshop on Benchmarking Multi-Target Tracking, 2016.
[3] Jonathon Luiten, A.O. & Leibe, B. HOTA: A Higher Order Metric for Evaluating Multi-Object Tracking. International Journal of Computer Vision, 2020.
[4] Li, Y., Huang, C. & Nevatia, R. Learning to associate: HybridBoosted multi-target tracker for crowded scene. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2009.



Contact