MOT17 Results

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

TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
ISDH_HDAv2
1. online method using public detections
54.5
±14.5
65.9
±10.8
75.2 622 (26.4)755 (32.1)46,693 207,093 63.3 88.4 2.6 3,010 (47.6)6,000 (94.8)3.6
MM-008988/ IEEE Transactions on Multimedia
Lif_T
2. using public detections
60.5
±13.0
65.6
±8.6
78.3 637 (27.0)791 (33.6)14,966 206,619 63.4 96.0 0.8 1,189 (18.8)3,476 (54.8)0.5
Anonymous submission
TT17
3. using public detections
54.9
±11.6
63.1
±8.6
77.2 575 (24.4)897 (38.1)20,236 233,295 58.7 94.2 1.1 1,088 (18.6)2,392 (40.8)2.5
Y. Zhang, H. Sheng, Y. Wu, S. Wang, W. Lyu, W. Ke, Z. Xiong. Long-term Tracking with Deep Tracklet Association. In IEEE Transactions on Image Processing, 2020.
LSST17
4. using public detections
54.7
±12.9
62.3
±9.5
75.9 480 (20.4)944 (40.1)26,091 228,434 59.5 92.8 1.5 1,243 (20.9)3,726 (62.6)1.5
Multi-Object Tracking with Multiple Cues and Switcher-Aware Classification
MPNTrack
5. using public detections
58.8
±12.4
61.7
±8.7
78.6 679 (28.8)788 (33.5)17,413 213,594 62.1 95.3 1.0 1,185 (19.1)2,265 (36.4)6.5
G. Brasó, L. Leal-Taixé. Learning a Neural Solver for Multiple Object Tracking. In CVPR, 2020.
CMT
6. using public detections
51.8
±12.7
60.7
±8.6
77.3 462 (19.6)1,009 (42.8)29,528 240,960 57.3 91.6 1.7 1,217 (21.2)2,008 (35.0)6.5
#Submission: TIP-21190-2019
ALBOD
7. online method using public detections
56.9
±12.1
58.7
±8.4
77.9 693 (29.4)633 (26.9)49,568 191,764 66.0 88.3 2.8 2,001 (30.3)3,163 (47.9)4.9
Anonymous submission
eTC17
8. using public detections
51.9
±12.9
58.1
±8.7
76.3 544 (23.1)836 (35.5)36,164 232,783 58.7 90.2 2.0 2,288 (38.9)3,071 (52.3)0.7
G. Wang, Y. Wang, H. Zhang, R. Gu, J. Hwang. Exploit the connectivity: Multi-object tracking with trackletnet. In Proceedings of the 27th ACM International Conference on Multimedia, 2019.
UnsupTrack
9. online method using public detections
61.7
±16.0
58.1
±11.1
78.3 640 (27.2)762 (32.4)16,872 197,632 65.0 95.6 1.0 1,864 (28.7)4,213 (64.8)2.0
Simple Unsupervised Multi-Object Tracking (Under Review)
TARCA
10. online method using public detections
55.9
±13.3
58.1
±9.1
77.5 571 (24.2)846 (35.9)20,141 227,151 59.7 94.4 1.1 1,784 (29.9)3,741 (62.6)6.9
Anonymous submission
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
VAN
11. online method using public detections
57.4
±12.8
57.9
±10.3
78.7 619 (26.3)794 (33.7)14,316 224,064 60.3 96.0 0.8 1,788 (29.7)2,586 (42.9)6.8
Anonymous submission
GMOT
12. using public detections
55.4
±12.2
57.9
±7.8
76.7 535 (22.7)817 (34.7)20,608 229,511 59.3 94.2 1.2 1,403 (23.7)2,765 (46.6)5.9
LXD, KHW @ HRI-SH
LSST17O
13. online method using public detections
52.7
±13.0
57.9
±9.3
76.2 421 (17.9)863 (36.6)22,512 241,936 57.1 93.5 1.3 2,167 (37.9)7,443 (130.3)1.8
Multi-Object Tracking with Multiple Cues and Switcher-Aware Classification
SAS_MOT17
14. using public detections
44.2
±0.0
57.2
±0.0
76.4 379 (16.1)1,044 (44.3)29,473 283,611 49.7 90.5 1.7 1,529 (30.7)2,644 (53.2)4.8
A. Maksai, P. Fua. Eliminating Exposure Bias and Metric Mismatch in Multiple Object Tracking. In CVPR, 2019.
TrajTrack
15. online method using public detections
56.0
±12.9
57.2
±10.5
78.6 532 (22.6)826 (35.1)14,378 231,212 59.0 95.9 0.8 2,546 (43.1)3,452 (58.5)1.4
Anonymous submission
ENFT17
16. using public detections
52.8
±13.1
57.1
±8.5
77.4 543 (23.1)867 (36.8)26,754 237,909 57.8 92.4 1.5 1,667 (28.8)2,557 (44.2)0.5
BUAA
TLMHT
17. using public detections
50.6
±12.5
56.5
±7.8
77.6 415 (17.6)1,022 (43.4)22,213 255,030 54.8 93.3 1.3 1,407 (25.7)2,079 (37.9)2.6
H. Sheng, J. Chen, Y. Zhang, W. Ke, Z. Xiong, J. Yu. Iterative Multiple Hypothesis Tracking with Tracklet-level Association. In IEEE Transactions on Circuits and Systems for Video Technology, 2018.
ISE_MOT17R
18. online method using public detections
60.1
±11.4
56.4
±9.2
78.2 672 (28.5)661 (28.1)23,168 199,483 64.6 94.0 1.3 2,556 (39.5)3,182 (49.2)7.2
MIFT
STRN_MOT17
19. online method using public detections
50.9
±11.6
56.0
±9.0
75.6 446 (18.9)797 (33.8)25,295 249,365 55.8 92.6 1.4 2,397 (43.0)9,363 (167.8)13.8
J. Xu, Y. Cao, Z. Zhang, H. Hu. Spatial-Temporal Relation Networks for Multi-Object Tracking. In ICCV, 2019.
TLO_bnw
20. using public detections
56.8
±13.6
56.0
±9.1
77.7 518 (22.0)872 (37.0)18,149 223,859 60.3 94.9 1.0 2,004 (33.2)3,567 (59.1)7.4
Anonymous submission
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
DMAN
21. online method using public detections
48.2
±12.3
55.7
±9.4
75.7 454 (19.3)902 (38.3)26,218 263,608 53.3 92.0 1.5 2,194 (41.2)5,378 (100.9)0.3
J. Zhu, H. Yang, N. Liu, M. Kim, W. Zhang, M. Yang. Online Multi-Object Tracking with Dual Matching Attention Networks. In ECCV, 2018.
Tracktor++v2
22. online method using public detections
56.3
±13.3
55.1
±10.8
78.8 498 (21.1)831 (35.3)8,866 235,449 58.3 97.4 0.5 1,987 (34.1)3,763 (64.6)1.5
P. Bergmann, T. Meinhardt, L. Leal-Taixé. Tracking without bells and whistles. In ICCV, 2019.
eHAF17
23. using public detections
51.8
±13.2
54.7
±9.1
77.0 551 (23.4)893 (37.9)33,212 236,772 58.0 90.8 1.9 1,834 (31.6)2,739 (47.2)0.7
H. Sheng, Y. Zhang, J. Chen, Z. Xiong, J. Zhang. Heterogeneous Association Graph Fusion for Target Association in Multiple Object Tracking. In IEEE Transactions on Circuits and Systems for Video Technology, 2018.
jCC
24. using public detections
51.2
±14.5
54.5
±10.9
75.9 493 (20.9)872 (37.0)25,937 247,822 56.1 92.4 1.5 1,802 (32.1)2,984 (53.2)1.8
M. Keuper, S. Tang, B. Andres, T. Brox, B. Schiele. Motion Segmentation amp; Multiple Object Tracking by Correlation Co-Clustering. In IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018.
NOTA
25. using public detections
51.3
±11.7
54.5
±7.6
76.7 403 (17.1)833 (35.4)20,148 252,531 55.2 93.9 1.1 2,285 (41.4)5,798 (105.0)17.8
L. Chen, H. Ai, R. Chen, Z. Zhuang. Aggregate Tracklet Appearance Features for Multi-Object Tracking. In IEEE Signal Processing Letters, 2019.
GNN_tracktor
26. online method using public detections
54.4
±12.9
54.1
±11.3
77.4 420 (17.8)880 (37.4)12,655 241,868 57.1 96.2 0.7 2,660 (46.6)3,991 (69.9)1.7
Anonymous submission
GMPHD_Rd17
27. online method using public detections
46.8
±14.7
54.1
±9.2
76.4 464 (19.7)784 (33.3)38,452 257,678 54.3 88.9 2.2 3,865 (71.1)8,097 (149.0)30.8
N. Baisa. Occlusion-robust Online Multi-object Visual Tracking using a GM-PHD Filter with a CNN-based Re-identification. In , 2019.
YOONKJ17
28. online method using public detections
51.4
±12.9
54.0
±9.5
77.0 500 (21.2)878 (37.3)29,051 243,202 56.9 91.7 1.6 2,118 (37.2)3,072 (54.0)3.4
K. YOON, J. GWAK, Y. SONG, Y. YOON, M. JEON. OneShotDA: Online Multi-object Tracker with One-shot-learning-based Data Association. In IEEE Access, 2020.
DAIST
29. online method using public detections
52.0
±12.5
53.9
±7.5
76.4 476 (20.2)787 (33.4)31,275 237,004 58.0 91.3 1.8 2,817 (48.6)5,875 (101.3)6.9
Anonymous submission
DAIST_
30. online method using public detections
52.1
±12.5
53.8
±7.5
76.5 463 (19.7)792 (33.6)29,931 237,913 57.8 91.6 1.7 2,689 (46.5)5,600 (96.8)6.9
Anonymous submission
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
TrctrD17
31. online method using public detections
53.7
±13.3
53.8
±10.6
77.2 458 (19.4)861 (36.6)11,731 247,447 56.1 96.4 0.7 1,947 (34.7)4,792 (85.4)4.9
Y. Xu, A. Osep, Y. Ban, R. Horaud, L. Leal-Taixe, X. Alameda-Pineda. How To Train Your Deep Multi-Object Tracker. In , 2019.
CRF_TRA
32. using public detections
53.1
±12.1
53.7
±9.3
76.1 571 (24.2)724 (30.7)27,194 234,991 58.4 92.4 1.5 2,518 (43.2)4,918 (84.3)1.4
Jun xiang, Chao Ma, Guohan Xu, Jianhua Hou, End-to-End Learning Deep CRF models for Multi-Object Tracking. In IEEE Transactions on Circuits and Systems for Video Technology, 2020
DEEP_TAMA
33. online method using public detections
50.3
±13.3
53.5
±10.1
76.7 453 (19.2)883 (37.5)25,479 252,996 55.2 92.4 1.4 2,192 (39.7)3,978 (72.1)1.5
Y. Yoon, D. Kim, K. Yoon, Y. Song, M. Jeon. Online Multiple Pedestrian Tracking using Deep Temporal Appearance Matching Association. In arXiv:1907.00831, 2019.
TLO_MHT
34. online method using public detections
53.3
±12.9
53.3
±9.5
76.5 471 (20.0)912 (38.7)22,161 238,959 57.6 93.6 1.2 2,434 (42.2)4,089 (70.9)2.0
Anonymous submission
MADA_NET
35. online method using public detections
57.6
±13.7
53.2
±11.0
78.5 523 (22.2)813 (34.5)10,475 224,948 60.1 97.0 0.6 3,657 (60.8)4,262 (70.9)1,233.1
Anonymous submission
MOTDT17
36. online method using public detections
50.9
±11.7
52.7
±7.5
76.6 413 (17.5)841 (35.7)24,069 250,768 55.6 92.9 1.4 2,474 (44.5)5,317 (95.7)18.3
C. Long, A. Haizhou, Z. Zijie, S. Chong. Real-time Multiple People Tracking with Deeply Learned Candidate Selection and Person Re-identification. In ICME, 2018.
Tracktor++
37. online method using public detections
53.5
±13.6
52.3
±10.6
78.0 459 (19.5)861 (36.6)12,201 248,047 56.0 96.3 0.7 2,072 (37.0)4,611 (82.3)1.5
P. Bergmann, T. Meinhardt, L. Leal-Taixé. Tracking without bells and whistles. In ICCV, 2019.
AM_ADM17
38. online method using public detections
48.1
±13.5
52.1
±10.7
76.7 316 (13.4)934 (39.7)25,061 265,495 52.9 92.3 1.4 2,214 (41.8)5,027 (94.9)5.7
S. Lee, M. Kim, S. Bae, Learning Discriminative Appearance Models for Online Multi-Object Tracking with Appearance Discriminability Measures, In IEEE Access, 2018.
track_bnw
39. online method using public detections
56.7
±13.4
52.1
±10.5
78.8 543 (23.1)813 (34.5)8,895 233,206 58.7 97.4 0.5 2,351 (40.1)3,155 (53.8)0.7
Anonymous submission
MHT_bLSTM
40. using public detections
47.5
±12.6
51.9
±9.8
77.5 429 (18.2)981 (41.7)25,981 268,042 52.5 91.9 1.5 2,069 (39.4)3,124 (59.5)1.9
C. Kim, F. Li, J. Rehg. Multi-object Tracking with Neural Gating Using Bilinear LSTM. In ECCV, 2018.
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
DASOT17
41. online method using public detections
49.5
±14.0
51.8
±10.6
76.9 481 (20.4)814 (34.6)33,640 247,370 56.2 90.4 1.9 4,142 (73.8)6,852 (122.0)9.1
Q. Chu, W. Ouyang, B. Liu, F. Zhu, N. Yu. DASOT: A Unified Framework Integrating Data Association and Single Object Tracking for Online Multi-Object Tracking. In Proceedings of the AAAI Conference on Artificial Intelligence, 2020.
msot
42. online method using public detections
59.2
±11.1
51.8
±9.2
78.0 630 (26.8)669 (28.4)18,968 208,569 63.0 94.9 1.1 2,648 (42.0)3,399 (53.9)0.9
Anonymous submission
UNS20
43. online method using public detections
51.3
±12.3
51.4
±10.0
76.3 432 (18.3)939 (39.9)13,599 259,389 54.0 95.7 0.8 1,872 (34.6)3,303 (61.1)12.2
Anonymous submission
DGCT
44. using public detections
54.5
±13.0
51.3
±10.7
79.0 495 (21.0)834 (35.4)10,471 243,143 56.9 96.8 0.6 2,865 (50.3)4,889 (85.9)7.0
CJY, HYW, KHW @ HRI-SH
TLO17
45. online method using public detections
52.6
±12.9
51.3
±9.7
76.6 460 (19.5)899 (38.2)20,089 244,930 56.6 94.1 1.1 2,530 (44.7)4,170 (73.7)25.2
Anonymous submission
EDMT17
46. using public detections
50.0
±13.9
51.3
±10.2
77.3 509 (21.6)855 (36.3)32,279 247,297 56.2 90.8 1.8 2,264 (40.3)3,260 (58.0)0.6
J. Chen, H. Sheng, Y. Zhang, Z. Xiong. Enhancing Detection Model for Multiple Hypothesis Tracking. In BMTT-PETS CVPRw, 2017.
HAM_SADF17
47. online method using public detections
48.3
±13.2
51.1
±10.3
77.2 402 (17.1)981 (41.7)20,967 269,038 52.3 93.4 1.2 1,871 (35.8)3,020 (57.7)5.0
Y. Yoon, A. Boragule, Y. Song, K. Yoon, M. Jeon. Online Multi-Object Tracking with Historical Appearance Matching and Scene Adaptive Detection Filtering. In IEEE AVSS, 2018.
JBNOT
48. using public detections
52.6
±12.3
50.8
±9.6
77.1 465 (19.7)844 (35.8)31,572 232,659 58.8 91.3 1.8 3,050 (51.9)3,792 (64.5)5.4
R. Henschel, Y. Zou, B. Rosenhahn. Multiple People Tracking using Body and Joint Detections. In CVPRW, 2019.
MOT_TBC
49. using public detections
53.9
±15.6
50.0
±10.2
76.8 476 (20.2)864 (36.7)24,584 232,670 58.8 93.1 1.4 2,945 (50.1)4,612 (78.5)6.7
Anonymous submission
PHD_GSDL17
50. online method using public detections
48.0
±13.6
49.6
±9.4
77.2 402 (17.1)838 (35.6)23,199 265,954 52.9 92.8 1.3 3,998 (75.6)8,886 (168.1)6.7
Z. Fu, P. Feng, F. Angelini, J. Chambers, S. Naqvi. Particle PHD Filter based Multiple Human Tracking using Online Group-Structured Dictionary Learning. In IEEE Access, 2018.
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
Seq2Seq
51. using public detections
52.7
±12.1
49.4
±9.4
77.0 417 (17.7)900 (38.2)10,819 253,890 55.0 96.6 0.6 2,396 (43.6)3,374 (61.3)2.6
Anonymous submission
MSRRT
52. online method using public detections
57.1
±11.1
49.3
±8.0
77.9 590 (25.1)790 (33.5)26,180 211,947 62.4 93.1 1.5 4,114 (65.9)4,595 (73.6)5.9
Anonymous submission
FAMNet
53. online method using public detections
52.0
±12.1
48.7
±9.9
76.5 450 (19.1)787 (33.4)14,138 253,616 55.1 95.6 0.8 3,072 (55.8)5,318 (96.6)0.0
P. Chu, H. Ling. FAMNet: Joint Learning of Feature, Affinity and Multi-dimensional Assignment for Online Multiple Object Tracking. In ICCV, 2019.
DAM_MOT
54. online method using public detections
47.0
±0.0
48.7
±0.0
76.9 397 (16.9)897 (38.1)28,933 267,896 52.5 91.1 1.6 2,140 (40.7)2,756 (52.5)18.7
Multi Object Tracking using Deep Structural Cost Minimization in Data Association
HDTR
55. using public detections
54.1
±11.5
48.4
±8.0
80.2 549 (23.3)819 (34.8)18,002 238,818 57.7 94.8 1.0 1,895 (32.9)2,693 (46.7)1.8
M. Babaee, A. Athar, G. Rigoll. Multiple People Tracking Using Hierarchical Deep Tracklet Re-identification. In arXiv preprint arXiv:1811.04091, 2018.
FPSN
56. online method using public detections
44.9
±13.8
48.4
±10.6
76.6 388 (16.5)844 (35.8)33,757 269,952 52.2 89.7 1.9 7,136 (136.8)14,491 (277.8)10.1
S. Lee, E. Kim. Multiple Object Tracking via Feature Pyramid Siamese Networks. In IEEE ACCESS, 2018.
OTCD_1
57. online method using public detections
48.6
±13.6
47.9
±11.1
76.9 382 (16.2)970 (41.2)18,499 268,204 52.5 94.1 1.0 3,502 (66.7)5,588 (106.5)15.5
Q. Liu, B. Liu, Y. Wu, W. Li, N. Yu. Real-Time Online Multi-Object Tracking in Compressed Domain. In IEEE Access, 2019.
FWT
58. using public detections
51.3
±13.2
47.6
±11.5
77.0 505 (21.4)830 (35.2)24,101 247,921 56.1 92.9 1.4 2,648 (47.2)4,279 (76.3)0.2
R. Henschel, L. Leal-Taixé, D. Cremers, B. Rosenhahn. Fusion of Head and Full-Body Detectors for Multi-Object Tracking. In Trajnet CVPRW, 2018.
MHT_DAM
59. using public detections
50.7
±13.7
47.2
±9.6
77.5 491 (20.8)869 (36.9)22,875 252,889 55.2 93.2 1.3 2,314 (41.9)2,865 (51.9)0.9
C. Kim, F. Li, A. Ciptadi, J. Rehg. Multiple Hypothesis Tracking Revisited. In ICCV, 2015.
GMPHDOGM17
60. online method using public detections
49.9
±13.3
47.1
±8.3
77.0 464 (19.7)895 (38.0)24,024 255,277 54.8 92.8 1.4 3,125 (57.1)3,540 (64.7)30.7
Y. Song, K. Yoon, Y. Yoon, K. Yow, M. Jeon. Online Multi-Object Tracking with GMPHD Filter and Occlusion Group Management. In IEEE Access, 2019.
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
AFN17
61. using public detections
51.5
±12.7
46.9
±8.3
77.6 485 (20.6)836 (35.5)22,391 248,420 56.0 93.4 1.3 2,593 (46.3)4,308 (77.0)1.8
H. Shen, L. Huang, C. Huang, W. Xu. Tracklet Association Tracker: An End-to-End Learning-based Association Approach for Multi-Object Tracking. In CoRR, 2018.
MASS
62. online method using public detections
46.9
±14.2
46.0
±10.7
76.1 399 (16.9)856 (36.3)25,733 269,116 52.3 92.0 1.4 4,478 (85.6)11,994 (229.3)17.1
H. Karunasekera, H. Wang, H. Zhang. Multiple Object Tracking With Attention to Appearance, Structure, Motion and Size. In IEEE Access, 2019.
MTDF17
63. online method using public detections
49.6
±13.7
45.2
±10.5
75.5 444 (18.9)779 (33.1)37,124 241,768 57.2 89.7 2.1 5,567 (97.4)9,260 (162.0)1.2
Z. Fu, F. Angelini, J. Chambers, S. Naqvi. Multi-Level Cooperative Fusion of GM-PHD Filters for Online Multiple Human Tracking. In IEEE Transactions on Multimedia, 2019.
MOCL
64. online method using public detections
49.5
±14.0
43.4
±11.2
77.2 487 (20.7)838 (35.6)25,373 254,131 55.0 92.4 1.4 5,164 (94.0)5,787 (105.3)148.0
ECCV-20/4696
LM_NN
65. using public detections
45.1
±13.3
43.2
±10.3
78.9 348 (14.8)1,088 (46.2)10,834 296,451 47.5 96.1 0.6 2,286 (48.2)2,463 (51.9)0.9
M. Babaee, Z. Li, G. Rigoll. A Dual CNN--RNN for Multiple People Tracking. In Neurocomputing, 2019.
PHD_GM
66. online method using public detections
48.8
±13.0
43.2
±9.3
76.7 449 (19.1)830 (35.2)26,260 257,971 54.3 92.1 1.5 4,407 (81.2)6,448 (118.8)22.3
R. Sanchez-Matilla, A. Cavallaro. A predictor of moving objects for First-Person vision. In Proceedings of IEEE International Conference Image Processing, 2019.
EAMTT
67. online method using public detections
42.6
±13.3
41.8
±9.7
76.0 300 (12.7)1,006 (42.7)30,711 288,474 48.9 90.0 1.7 4,488 (91.8)5,720 (117.0)12.0
R. Sanchez-Matilla, F. Poiesi, A. Cavallaro. Online Multi-target Tracking with Strong and Weak Detections. In Computer Vision -- ECCV 2016 Workshops, 2016.
EDA_GNN
68. online method using public detections
45.5
±13.8
40.5
±10.2
76.3 368 (15.6)955 (40.6)25,685 277,663 50.8 91.8 1.4 4,091 (80.5)5,579 (109.8)39.3
Paper ID 2713
HISP_DAL17
69. online method using public detections
45.4
±13.9
39.9
±9.5
77.3 349 (14.8)922 (39.2)21,820 277,473 50.8 92.9 1.2 8,727 (171.7)7,147 (140.6)3.2
N. Baisa. Robust Online Multi-target Visual Tracking using a HISP Filter with Discriminative Deep Appearance Learning. In CoRR, 2019.
SORT17
70. online method using public detections
43.1
±13.3
39.8
±10.3
77.8 295 (12.5)997 (42.3)28,398 287,582 49.0 90.7 1.6 4,852 (99.0)7,127 (145.4)143.3
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.
TrackerMOTAIDF1MOTPMTMLFPFNRecallPrecisionFAFID Sw.FragHz
IOU17
71. using public detections
45.5
±13.6
39.4
±10.3
76.9 369 (15.7)953 (40.5)19,993 281,643 50.1 93.4 1.1 5,988 (119.6)7,404 (147.8)1,522.9
E. Bochinski, V. Eiselein, T. Sikora. High-Speed Tracking-by-Detection Without Using Image Information. In International Workshop on Traffic and Street Surveillance for Safety and Security at IEEE AVSS 2017, 2017.
Tracker
72. online method using public detections
42.4
±13.9
39.2
±10.5
76.5 272 (11.5)1,045 (44.4)21,614 298,128 47.2 92.5 1.2 5,081 (107.7)6,259 (132.7)66.9
Anonymous submission
GMPHD_SHA
73. online method using public detections
43.7
±0.0
39.2
±0.0
76.5 276 (11.7)1,012 (43.0)25,935 287,758 49.0 91.4 1.5 3,838 (78.3)5,056 (103.2)9.2
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.
HISP_T17
74. online method using public detections
44.6
±0.0
38.8
±0.0
77.2 355 (15.1)913 (38.8)25,478 276,395 51.0 91.9 1.4 10,617 (208.1)7,487 (146.8)4.7
N. Baisa. Online Multi-target Visual Tracking using a HISP Filter. In Proceedings of the 13th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISAPP,, 2018.
MST
75. online method using public detections
39.2
±19.4
37.6
±10.7
75.9 477 (20.3)730 (31.0)89,414 245,273 56.5 78.1 5.0 8,532 (150.9)9,606 (169.9)370.9
GMPHD_KCF
76. online method using public detections
39.6
±13.0
36.6
±7.6
74.5 208 (8.8)1,019 (43.3)50,903 284,228 49.6 84.6 2.9 5,811 (117.1)7,414 (149.4)3.3
T. Kutschbach, E. Bochinski, V. Eiselein, T. Sikora. Sequential Sensor Fusion Combining Probability Hypothesis Density and Kernelized Correlation Filters for Multi-Object Tracking in Video Data. In International Workshop on Traffic and Street Surveillance for Safety and Security at IEEE AVSS 2017, 2017.
GMPHD_DAL
77. online method using public detections
44.4
±13.9
36.2
±9.2
77.4 350 (14.9)927 (39.4)19,170 283,380 49.8 93.6 1.1 11,137 (223.7)13,900 (279.3)3.4
N. Baisa. Online Multi-object Visual Tracking using a GM-PHD Filter with Deep Appearance Learning. In 2019 22th International Conference on Information Fusion (FUSION), 2019.
GM_PHD
78. online method using public detections
36.4
±14.0
33.9
±9.6
76.2 97 (4.1)1,349 (57.3)23,723 330,767 41.4 90.8 1.3 4,607 (111.3)11,317 (273.5)38.4
V. Eiselein, D. Arp, M. Pätzold, T. Sikora. Real-time Multi-Human Tracking using a Probability Hypothesis Density Filter and multiple detectors. In 9th IEEE International Conference on Advanced Video and Signal-Based Surveillance, 2012.
GMPHD_N1Tr
79. online method using public detections
42.1
±13.0
33.9
±9.5
77.7 280 (11.9)1,005 (42.7)18,214 297,646 47.2 93.6 1.0 10,698 (226.4)10,864 (229.9)9.9
N. Baisa, A. Wallace. Development of a N-type GM-PHD filter for multiple target, multiple type visual tracking. In Journal of Visual Communication and Image Representation, 2019.
SequencesFramesTrajectoriesBoxes
21177572355564228

Difficulty Analysis

Sequence difficulty (from easiest to hardest, measured by average MOTA)

MOT17-03-SDP

MOT17-03-SDP

(73.6 MOTA)

MOT17-03-FRCNN

MOT17-03-FRCNN

(61.8 MOTA)

MOT17-03-DPM

MOT17-03-DPM

(55.4 MOTA)

...

...

MOT17-14-DPM

MOT17-14-DPM

(25.5 MOTA)

MOT17-14-FRCNN

MOT17-14-FRCNN

(24.3 MOTA)


Evaluation Measures

Lower is better. Higher is better.
Measure Better Perfect Description
MOTA higher 100 % Multiple Object Tracking Accuracy [1]. This measure combines three error sources: false positives, missed targets and identity switches.
MOTP higher 100 % Multiple Object Tracking Precision [1]. The misalignment between the annotated and the predicted bounding boxes.
IDF1 higher 100 % ID F1 Score [2]. The ratio of correctly identified detections over the average number of ground-truth and computed detections.
FAF lower 0 The average number of false alarms per frame.
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 0 The total number of false positives.
FN lower 0 The total number of false negatives (missed targets).
ID Sw. lower 0 The total number of identity switches. Please note that we follow the stricter definition of identity switches as described in [3].
Frag lower 0 The 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 public detections This method used the provided detection set as input.
new This entry has been submitted or updated less than a week ago.
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] 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.