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 RankMOTAMOTPFAFMTMLFPFNID Sw.FragHzDetector
AMIR
1. online method using public detections
17.5
47.2
±7.7
75.80.514.0% 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.
CDA_DDALv2
2. online method using public detections
28.0
43.9
±7.8
74.71.110.7% 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.
CEM
3. using public detections
29.4
33.2
±7.9
75.81.27.8% 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.
CMRZF
4. using public detections
26.1
30.4
±10.8
77.80.22.9% 70.5% 1,421124,4831,030 (32.5)733 (23.1)16.9Public
Anonymous submission
DCCRF16
5. online method using public detections
23.2
44.8
±9.5
75.60.914.1% 42.3% 5,61394,125968 (20.0)1,378 (28.5)0.1Public
Anonymous submission
DeepAC
6. online method using public detections
28.9
38.8
±9.3
74.90.99.1% 42.8% 5,444103,1742,886 (66.5)6,592 (151.9)21.1Public
Anonymous submission
dmot
7. using public detections
27.9
40.7
±8.3
75.21.612.0% 44.1% 9,31997,992773 (16.7)1,106 (23.9)6.6Public
Anonymous submission
DP_NMS
8. using public detections
24.2
32.2
±9.8
76.40.25.4% 62.1% 1,123121,579972 (29.2)944 (28.3)212.6Public
H. Pirsiavash, D. Ramanan, C. Fowlkes. Globally-Optimal Greedy Algorithms for Tracking a Variable Number of Objects. In CVPR, 2011.
DQNTracker
9. online method using public detections
31.7
33.7
±13.7
75.40.96.9% 59.3% 5,210113,8651,744 (46.4)4,184 (111.4)9.9Public
Anonymous submission
EAGS16
10. using public detections
15.3
47.4
±10.4
75.91.417.3% 42.7% 8,36986,931575 (11.0)913 (17.5)197.3Public
#MM-007925 Enhancing Association Graph with Super-voxel for Multi-target Tracking
TrackerAvg RankMOTAMOTPFAFMTMLFPFNID Sw.FragHzDetector
EAMTT_pub
11. online method using public detections
30.9
38.8
±8.5
75.11.47.9% 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
EDMT
12. using public detections
20.1
45.3
±9.1
75.91.917.0% 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.
EMOT
13. online method using public detections
32.8
35.7
±7.9
73.42.811.6% 43.7% 16,56699,794826 (18.2)2,123 (46.9)5,919.0Public
Anonymous submission
ERCTracker
14. online method using public detections
24.3
32.3
±9.4
76.40.25.7% 62.1% 1,193121,333953 (28.5)943 (28.2)32.0Public
Anonymous submission
FWT
15. using public detections
20.2
47.8
±9.4
75.51.519.1% 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.
GMCSS
16. online method using public detections
35.6
38.3
±9.0
74.32.89.4% 46.6% 16,49195,303735 (15.4)2,122 (44.5)0.4Public
Anonymous submission
GMPHD_AM
17. online method using public detections
31.2
30.6
±6.7
74.80.85.9% 53.1% 4,982120,698930 (27.5)1,856 (54.9)7.9Public
Anonymous submission
GMPHD_HDA
18. online method using public detections
26.6
30.5
±6.9
75.40.94.6% 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.
GM_PHD_N1T
19. online method using public detections
30.5
31.6
±8.6
76.60.85.5% 55.2% 4,767115,6454,348 (118.9)3,986 (109.0)9.9Public
Anonymous submission
HAF16
20. using public detections
21.8
45.7
±8.9
76.01.715.4% 41.4% 10,03888,319660 (12.8)985 (19.1)0.7Public
Anonymous submission
TrackerAvg RankMOTAMOTPFAFMTMLFPFNID Sw.FragHzDetector
HCC
21. using public detections
9.0
49.3
±10.2
79.00.917.8% 39.9% 5,33386,795391 (7.5)535 (10.2)0.8Public
Anonymous submission
HFCLP
22. using public detections
24.2
42.7
±7.6
76.31.412.9% 40.2% 8,50294,2661,676 (34.7)1,792 (37.1)19.7Public
Anonymous submission
JCSTD
23. online method using public detections
23.9
47.4
±8.3
74.41.414.4% 36.4% 8,07786,6311,266 (24.1)2,696 (51.4)3.3Public
Anonymous submission
JMC
24. using public detections
17.6
46.3
±9.0
75.71.115.5% 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.
JPDA_m
25. using public detections
24.1
26.2
±6.1
76.30.64.1% 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.
LINF1
26. using public detections
26.2
41.0
±9.5
74.81.311.6% 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.
LKDeep
27. online method using public detections
34.2
31.8
±19.3
74.81.06.2% 53.5% 6,179115,8012,389 (65.5)5,745 (157.5)32.0Public
Anonymous submission
LMP
28. using public detections
12.2
48.8
±9.8
79.01.118.2% 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.
LP2D
29. using public detections
24.1
35.7
±10.1
75.80.98.7% 50.7% 5,084111,163915 (23.4)1,264 (32.4)49.3Public
MOT baseline: Linear programming on 2D image coordinates.
LTTSC-CRF
30. using public detections
30.8
37.6
±9.9
75.92.09.6% 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.
TrackerAvg RankMOTAMOTPFAFMTMLFPFNID Sw.FragHzDetector
MCjoint
31. using public detections
16.1
47.1
±10.8
76.31.120.4% 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.
MHT_DAM
32. using public detections
17.4
45.8
±8.9
76.31.116.2% 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.
MOT_M_hun
33. using public detections
25.3
39.0
±10.3
75.52.613.7% 40.1% 15,34595,029843 (17.6)1,790 (37.4)5,919.0Public
Anonymous submission
NHL
34. using public detections
29.5
45.1
±8.5
72.52.115.9% 37.3% 12,60585,6911,747 (33.0)2,033 (38.4)0.3Public
Anonymous submission
NLLMPa
35. using public detections
11.7
47.6
±10.6
78.51.017.0% 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.
NOMT
36. using public detections
14.4
46.4
±9.9
76.61.618.3% 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.
oBot
37. online method using public detections
28.1
42.5
±20.4
75.21.812.6% 40.7% 10,42092,8921,559 (31.8)1,639 (33.4)2.3Public
Anonymous BMVC submission
oICF
38. online method using public detections
27.2
43.2
±10.2
74.31.111.3% 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.
OVBT
39. online method using public detections
36.2
38.4
±8.8
75.41.97.5% 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, .
PAOT
40. online method using public detections
27.8
31.5
±9.0
77.30.54.3% 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.
TrackerAvg RankMOTAMOTPFAFMTMLFPFNID Sw.FragHzDetector
PRMOT
41. using public detections
26.9
38.4
±8.9
75.42.712.4% 47.3% 15,76495,796741 (15.6)885 (18.6)2,959.5Public
Anonymous submission
QuadMOT16
42. using public detections
20.2
44.1
±9.4
76.41.114.6% 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.
RAR16pub
43. online method using public detections
24.3
45.9
±9.7
74.81.213.2% 41.9% 6,87191,173648 (13.0)1,992 (39.8)0.9Public
Anonymous ICCV submission
rookie_ksp
44. using public detections
32.5
24.8
±7.7
76.40.22.4% 66.1% 1,421132,3613,343 (122.0)4,886 (178.3)19.7Public
Anonymous submission
SLT
45. online method using public detections
27.7
41.5
±10.4
75.31.411.7% 37.0% 8,07496,9561,705 (36.4)3,170 (67.7)9.6Public
Anonymous submission
SMOT
46. using public detections
43.5
29.7
±7.3
75.22.95.3% 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.
STAM16
47. online method using public detections
24.0
46.0
±9.1
74.91.214.6% 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.
STMOT
48. using public detections
32.2
38.2
±8.6
74.02.412.9% 42.8% 13,97097,3841,342 (28.8)2,828 (60.7)5,919.0Public
Anonymous submission
TBD
49. using public detections
30.8
33.7
±9.2
76.51.07.2% 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.
tMOT
50. using public detections
26.7
28.9
±10.5
75.10.65.8% 63.0% 3,754125,494468 (15.0)694 (22.3)11.8Public
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

(51.6% MOTA)

MOT16-06

MOT16-06

(44.1% MOTA)

MOT16-07

MOT16-07

(38.6% MOTA)

...

...

MOT16-08

MOT16-08

(29.4% MOTA)

MOT16-14

MOT16-14

(24.3% 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.
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 [2].
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] 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.