MOT16 Results

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


Showing only entries that use public detections!

TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
rookie_ksp
1. using public detections
39.1
24.8
±7.7
11.52.4% 66.1% 1,421132,3613,343 (122.0)4,886 (178.3)19.7Public
Anonymous submission
DP_NMS
2. using public detections
27.1
26.2
±9.3
31.24.1% 67.5% 3,689130,557365 (12.9)638 (22.5)5.9Public
H. Pirsiavash, D. Ramanan, C. Fowlkes. Globally-Optimal Greedy Algorithms for Tracking a Variable Number of Objects. In CVPR, 2011.
JPDA_m
3. using public detections
28.3
26.2
±6.1
0.04.1% 67.5% 3,689130,549365 (12.9)638 (22.5)22.2Public
H. Rezatofighi, A. Milan, Z. Zhang, Q. Shi, A. Dick, I. Reid. Joint Probabilistic Data Association Revisited. In ICCV, 2015.
tMOT
4. using public detections
30.1
28.9
±10.5
32.45.8% 63.0% 3,754125,494468 (15.0)694 (22.3)11.8Public
Anonymous submission
SMOT
5. using public detections
44.1
29.7
±7.3
0.05.3% 47.7% 17,426107,5523,108 (75.8)4,483 (109.3)0.2Public
C. Dicle, O. Camps, M. Sznaier. The Way They Move: Tracking Targets with Similar Appearance. In ICCV, 2013.
CMRZF
6. using public detections
29.3
30.4
±10.8
29.42.9% 70.5% 1,421124,4831,030 (32.5)733 (23.1)16.9Public
Anonymous submission
GMPHD_HDA
7. online method using public detections
27.1
30.5
±6.9
33.44.6% 59.7% 5,169120,970539 (16.0)731 (21.7)13.6Public
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.
GMPHD_AM
8. online method using public detections
31.1
30.6
±6.7
30.25.9% 53.1% 4,982120,698930 (27.5)1,856 (54.9)7.9Public
Anonymous submission
CppSORT
9. online method using public detections
31.1
31.5
±9.0
27.74.3% 59.9% 3,048120,2781,587 (46.6)2,239 (65.8)687.1Public
S. Murray. Real-Time Multiple Object Tracking - A Study on the Importance of Speed. In arXiv preprint arXiv:1709.03572, 2017.
GM_PHD_N1T
10. online method using public detections
34.8
31.6
±8.6
19.75.5% 55.2% 4,767115,6454,348 (118.9)3,986 (109.0)9.9Public
Anonymous submission
TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
LKDeep
11. online method using public detections
34.6
31.8
±19.3
27.66.2% 53.5% 6,179115,8012,389 (65.5)5,745 (157.5)32.0Public
Anonymous submission
ERCTracker
12. online method using public detections
28.4
32.3
±9.4
29.25.7% 62.1% 1,193121,333953 (28.5)943 (28.2)32.0Public
Anonymous submission
CEM
13. using public detections
29.8
33.2
±7.9
0.07.8% 54.4% 6,837114,322642 (17.2)731 (19.6)0.3Public
A. Milan, S. Roth, K. Schindler. Continuous Energy Minimization for Multitarget Tracking. In IEEE TPAMI, 2014.
DQNTracker
14. online method using public detections
32.3
33.7
±13.7
31.76.9% 59.3% 5,210113,8651,744 (46.4)4,184 (111.4)9.9Public
Anonymous submission
TBD
15. using public detections
37.6
33.7
±9.2
0.07.2% 54.2% 5,804112,5872,418 (63.2)2,252 (58.9)1.3Public
A. Geiger, M. Lauer, C. Wojek, C. Stiller, R. Urtasun. 3D Traffic Scene Understanding from Movable Platforms. In Pattern Analysis and Machine Intelligence (PAMI), 2014.
EMOT
16. online method using public detections
27.1
35.7
±7.9
45.611.6% 43.7% 16,56699,794826 (18.2)2,123 (46.9)5,919.0Public
Anonymous submission
LP2D
17. using public detections
27.1
35.7
±10.1
34.28.7% 50.7% 5,084111,163915 (23.4)1,264 (32.4)49.3Public
MOT baseline: Linear programming on 2D image coordinates.
DMMOT
18. online method using public detections
22.5
37.3
±11.1
52.613.6% 44.5% 20,88692,926451 (9.2)1,799 (36.7)0.5Public
Anonymous submission
LTTSC-CRF
19. using public detections
29.5
37.6
±9.9
42.19.6% 55.2% 11,969101,343481 (10.8)1,012 (22.8)0.6Public
N. Le, A. Heili, M. Odobez. Long-Term Time-Sensitive Costs for CRF-Based Tracking by Detection. In ECCVw, 2016.
STMOT
20. using public detections
30.6
38.2
±8.6
44.512.9% 42.8% 13,97097,3841,342 (28.8)2,828 (60.7)5,919.0Public
Anonymous submission
TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
GMCSS
21. online method using public detections
30.2
38.3
±9.0
44.79.4% 46.6% 16,49195,303735 (15.4)2,122 (44.5)0.4Public
Anonymous submission
OVBT
22. online method using public detections
36.5
38.4
±8.8
37.87.5% 47.3% 11,51799,4631,321 (29.1)2,140 (47.1)0.3Public
Y. Ban, S. Ba, X. Alameda-Pineda, R. Horaud. Tracking Multiple Persons Based on a Variational Bayesian Model. In BMTT 2016, .
PRMOT
23. using public detections
27.0
38.4
±8.9
39.912.4% 47.3% 15,76495,796741 (15.6)885 (18.6)2,959.5Public
Anonymous submission
EAMTT_pub
24. online method using public detections
27.6
38.8
±8.5
42.47.9% 49.1% 8,114102,452965 (22.0)1,657 (37.8)11.8Public
R. Sanchez-Matilla, F. Poiesi, A. Cavallaro "Multi-target tracking with strong and weak detections" in BMTT ECCVw 2016
DeepAC
25. online method using public detections
27.8
38.8
±9.3
33.19.1% 42.8% 5,444103,1742,886 (66.5)6,592 (151.9)21.1Public
Anonymous submission
MOT_M_hun
26. using public detections
22.3
39.0
±10.3
47.513.7% 40.1% 15,34595,029843 (17.6)1,790 (37.4)5,919.0Public
Anonymous submission
ONEEC
27. online method using public detections
30.4
39.3
±9.0
49.110.9% 46.5% 12,52497,451618 (13.3)2,336 (50.2)0.3Public
Anonymous submission
STFP
28. online method using public detections
30.4
39.8
±8.9
47.413.0% 41.4% 12,11896,755950 (20.2)2,630 (56.0)0.4Public
Anonymous submission
dmot
29. using public detections
24.4
40.7
±8.3
40.612.0% 44.1% 9,31997,992773 (16.7)1,106 (23.9)6.6Public
Anonymous submission
STbase
30. online method using public detections
29.5
41.0
±8.7
48.013.0% 41.4% 9,85696,799943 (20.1)2,595 (55.3)0.4Public
Anonymous submission
TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
LINF1
31. using public detections
23.3
41.0
±9.5
45.711.6% 51.3% 7,89699,224430 (9.4)963 (21.1)4.2Public
L. Fagot-Bouquet, R. Audigier, Y. Dhome, F. Lerasle. Improving Multi-Frame Data Association with Sparse Representations for Robust Near-Online Multi-Object Tracking. In ECCV, 2016.
SLT
32. online method using public detections
28.3
41.5
±10.4
41.611.7% 37.0% 8,07496,9561,705 (36.4)3,170 (67.7)9.6Public
Anonymous submission
OVMOT
33. online method using public detections
24.8
41.9
±8.8
50.115.0% 43.0% 10,71294,510626 (13.0)2,008 (41.7)0.4Public
Anonymous submission
oBot
34. online method using public detections
25.8
42.5
±20.4
40.812.6% 40.7% 10,42092,8921,559 (31.8)1,639 (33.4)2.3Public
Anonymous BMVC submission
HFCLP
35. using public detections
23.6
42.7
±7.6
35.112.9% 40.2% 8,50294,2661,676 (34.7)1,792 (37.1)19.7Public
Anonymous submission
oICF
36. online method using public detections
24.3
43.2
±10.2
49.311.3% 48.5% 6,65196,515381 (8.1)1,404 (29.8)0.4Public
H. Kieritz, S. Becker, W. Hübner, M. Arens. Online Multi-Person Tracking using Integral Channel Features. In IEEE Advanced Video and Signal-based Surveillance (AVSS) 2016, 2016.
ASSMOT
37. using public detections
21.8
43.2
±9.3
50.114.4% 44.5% 7,41395,458666 (14.0)2,010 (42.2)0.4Public
Anonymous submission
RMFP
38. online method using public detections
24.9
43.4
±9.2
50.414.9% 44.0% 7,55995,015682 (14.2)1,999 (41.7)0.3Public
Anonymous submission
NOSVM
39. using public detections new
23.3
43.6
±9.4
51.115.3% 42.8% 9,10692,991718 (14.7)2,084 (42.5)0.4Public
Anonymous submission
overMOT
40. online method using public detections new
22.8
43.7
±9.3
50.815.2% 43.0% 8,89193,036662 (13.5)1,844 (37.7)0.4Public
Anonymous submission
TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
CDA_DDALv2
41. online method using public detections
22.5
43.9
±7.8
45.110.7% 44.4% 6,45095,175676 (14.1)1,795 (37.6)0.5Public
S. Bae and K. Yoon, Confidence-Based Data Association and Discriminative Deep Appearance Learning for Robust Online Multi-Object Tracking, In IEEE TPAMI, 2017.
QuadMOT16
42. using public detections
23.4
44.1
±9.4
38.314.6% 44.9% 6,38894,775745 (15.5)1,096 (22.8)1.8Public
J. Son, M. Baek, M. Cho, B. Han. Multi-Object Tracking with Quadruplet Convolutional Neural Networks. In CVPR, 2017.
DCCRF16
43. online method using public detections
22.1
44.8
±9.5
39.714.1% 42.3% 5,61394,125968 (20.0)1,378 (28.5)0.1Public
Anonymous submission
NHL
44. using public detections
26.3
45.1
±8.5
32.315.9% 37.3% 12,60585,6911,747 (33.0)2,033 (38.4)0.3Public
Anonymous submission
EDMT
45. using public detections
15.2
45.3
±9.1
47.917.0% 39.9% 11,12287,890639 (12.3)946 (18.3)1.8Public
J. Chen, H. Sheng, Y. Zhang, Z. Xiong. Enhancing Detection Model for Multiple Hypothesis Tracking. In BMTT-PETS CVPRw, 2017.
HAF16
46. using public detections
17.9
45.7
±8.9
47.215.4% 41.4% 10,03888,319660 (12.8)985 (19.1)0.7Public
Anonymous submission
MHT_DAM
47. using public detections
17.2
45.8
±8.9
46.116.2% 43.2% 6,41291,758590 (11.9)781 (15.7)0.8Public
C. Kim, F. Li, A. Ciptadi, J. Rehg. Multiple Hypothesis Tracking Revisited. In ICCV, 2015.
RAR16pub
48. online method using public detections
21.0
45.9
±9.7
48.813.2% 41.9% 6,87191,173648 (13.0)1,992 (39.8)0.9Public
Anonymous ICCV submission
STAM16
49. online method using public detections
21.5
46.0
±9.1
50.014.6% 43.6% 6,89591,117473 (9.5)1,422 (28.4)0.2Public
Q. Chu, W. Ouyang, H. Li, X. Wang, B. Liu, N. Yu. Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism. In arXiv preprint arXiv:1708.02843, 2017.
JMC
50. using public detections
15.4
46.3
±9.0
46.315.5% 39.7% 6,37390,914657 (13.1)1,114 (22.2)0.8Public
S. Tang, B. Andres, M. Andriluka, B. Schiele. Multi-Person Tracking by Multicuts and Deep Matching. In BMTT, 2016.
TrackerAvg Rank MOTAIDF1MTMLFPFNID Sw.FragHzDetector
NOMT
51. using public detections
11.9
46.4
±9.9
53.318.3% 41.4% 9,75387,565359 (6.9)504 (9.7)2.6Public
W. Choi. Near-Online Multi-target Tracking with Aggregated Local Flow Descriptor. In ICCV, 2015.
MCjoint
52. using public detections
13.8
47.1
±10.8
52.320.4% 46.9% 6,70389,368370 (7.3)598 (11.7)0.6Public
M. Keuper, S. Tang, Z. Yu, B. Andres, T. Brox, B. Schiele. A Multi-cut Formulation for Joint Segmentation and Tracking of Multiple Objects. In CoRR, 2016.
AMIR
53. online method using public detections
15.4
47.2
±7.7
46.314.0% 41.6% 2,68192,856774 (15.8)1,675 (34.1)1.0Public
A. Sadeghian, A. Alahi, S. Savarese. Tracking The Untrackable: Learning To Track Multiple Cues with Long-Term Dependencies. In arXiv preprint arXiv:1701.01909, 2017.
JCSTD
54. online method using public detections
20.5
47.4
±8.3
41.114.4% 36.4% 8,07786,6311,266 (24.1)2,696 (51.4)3.3Public
Anonymous submission
EAGS16
55. using public detections
10.8
47.4
±10.4
50.117.3% 42.7% 8,36986,931575 (11.0)913 (17.5)197.3Public
Enhancing Association Graph with Super-voxel for Multi-target Tracking
NLLMPa
56. using public detections
13.2
47.6
±10.6
47.317.0% 40.4% 5,84489,093629 (12.3)768 (15.0)8.3Public
E. Levinkov, J. Uhrig, S. Tang, M. Omran, E. Insafutdinov, A. Kirillov, C. Rother, T. Brox, B. Schiele, B. Andres. Joint Graph Decomposition and Node Labeling: Problem, Algorithms, Applications. In CVPR, 2017.
FWT
57. using public detections
16.5
47.8
±9.4
44.319.1% 38.2% 8,88685,487852 (16.0)1,534 (28.9)0.6Public
R. Henschel, L. Leal-Taixé, D. Cremers, B. Rosenhahn. A Novel Multi-Detector Fusion Framework for Multi-Object Tracking. In arXiv preprint arXiv:1705.08314, 2017.
LMP
58. using public detections
11.1
48.8
±9.8
51.318.2% 40.1% 6,65486,245481 (9.1)595 (11.3)0.5Public
S. Tang, M. Andriluka, B. Andres, B. Schiele. Multiple People Tracking with Lifted Multicut and Person Re-identification. In CVPR, 2017.
HCC
59. using public detections
8.8
49.3
±10.2
50.717.8% 39.9% 5,33386,795391 (7.5)535 (10.2)0.8Public
Anonymous submission

Due to a minor bug in the export script, all results were re-evaluated on April 11, 2016. Here is the old snapshot of the leaderboard.


Benchmark Statistics

SequencesFramesTrajectoriesBoxes
75919759182326

Difficulty Analysis

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

MOT16-03

MOT16-03

(51.2% MOTA)

MOT16-06

MOT16-06

(44.0% MOTA)

MOT16-07

MOT16-07

(38.0% MOTA)

...

...

MOT16-08

MOT16-08

(29.2% MOTA)

MOT16-14

MOT16-14

(24.0% MOTA)


Evaluation Measures

Lower is better. Higher is better.
Measure Better Perfect Description
Avg Rank lower 1 This is the rank of each tracker averaged over all present evaluation measures.
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.

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.
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.