MEEM: Robust Tracking via Multiple Experts using
Entropy Minimization
Jianming Zhang, Shugao Ma, Stan Sclaroff
Department of Computer Science, Boston University, USA
We provide in this document the following statistics of MEEM and other trackers
on the benchmark dataset of [1] in Experiment I and our newly collected sequences in
Experiment II.
1. Fig. 1 (Experiment I): the average precision plots and the success plots for the compared trackers (LSHT [2], LSST [3] and SPLTT [4]), which are not included in [1].
For the other 29 compared trackers, readers are referred to [1].
2. Fig. 2 (Experiment I): the average precision plot ranking scores and the AUC ranking scores of the five leading trackers on different subsets of test sequences.
3. Fig. 3 and 4 (Experiment I): the precision plot ranking score tables and the AUC
ranking score tables for the top five trackers.
4. Fig. 5 and 6 (Experiment I): the precision plot ranking score tables and the AUC
ranking score tables for MEEM and its baselines.
5. Fig. 7 (Experiment II): the precision plot ranking score table and the AUC ranking score table for MEEM and the compared algorithms on our newly collected
sequences.
Detailed explanations are included in the caption of each figure. For more information
about the test sequences, evaluation protocols, and the 29 trackers evaluated in [1],
readers are directed to [1] and its project website.
References
1. Wu, Y., Lim, J., Yang, M.H.: Online object tracking: A benchmark. In: CVPR. (2013)
2. He, S., Yang, Q., Lau, R.W., Wang, J., Yang, M.H.: Visual tracking via locality sensitive
histograms. In: CVPR. (2013)
3. Wang, D., Lu, H., Yang, M.H.: Least soft-thresold squares tracking. In: CVPR. (2013)
4. Supancic III, J.S., Ramanan, D.: Self-paced learning for long-term tracking. In: CVPR.
(2013)
5. Hare, S., Saffari, A., Torr, P.H.: Struck: Structured output tracking with kernels. In: ICCV.
(2011)
6. Zhong, W., Lu, H., Yang, M.H.: Robust object tracking via sparsity-based collaborative
model. In: CVPR. (2012)
7. Kalal, Z., Matas, J., Mikolajczyk, K.: Pn learning: Bootstrapping binary classifiers by structural constraints. In: CVPR. (2010)
8. Kwon, J., Lee, K.M.: Visual tracking decomposition. In: CVPR. (2010)
9. Kwon, J., Lee, K.M.: Tracking by sampling trackers. In: ICCV. (2011)
10. Jia, X., Lu, H., Yang, M.H.: Visual tracking via adaptive structural local sparse appearance
model. In: CVPR. (2012)
11. Dinh, T.B., Vo, N., Medioni, G.: Context tracker: Exploring supporters and distracters in
unconstrained environments. In: CVPR. (2011)
2
Supplementary Materials
Precision plots of OPE
Precision plots of TRE
1
0.8
0.8
0.8
0.7
0.7
0.7
0.6
0.5
0.4
MEEM [0.840]
SPLTT [0.725]
LSHT [0.605]
LSST [0.514]
0.3
0.1
0
0
1
10
20
30
40
Location error threshold
0.6
0.5
0.4
0.3
MEEM [0.832]
LSHT [0.606]
LSST [0.567]
0.2
0.1
0
0
50
Success plots of OPE
Precision
0.9
0.2
1
10
20
30
40
Location error threshold
0.5
0.4
0.3
0.1
0
0
50
Success plots of TRE
1
0.8
0.8
0.8
0.5
0.4
0.3
0.2
0.1
0
0
MEEM [0.572]
SPLTT [0.523]
LSHT [0.421]
LSST [0.354]
0.2
0.4
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.6
0.8
Overlap threshold
1
Success rate
0.9
0.6
0
0
MEEM [0.585]
LSHT [0.433]
LSST [0.419]
0.2
0.4
0.6
0.8
1
10
20
30
40
Location error threshold
50
Success plots of SRE
0.7
0.6
0.5
0.4
0.3
0.2
0.1
Overlap threshold
MEEM [0.769]
LSHT [0.543]
LSST [0.486]
0.2
0.9
0.7
Precision plots of SRE
0.6
0.9
Success rate
Success rate
1
0.9
Precision
Precision
1
0.9
0
0
MEEM [0.518]
LSHT [0.371]
LSST [0.329]
0.2
0.4
0.6
0.8
Overlap threshold
1
Fig. 1. Average precision plots (top row) and success plots (bottom row) of MEEM and three
recent trackers not included in [1] (LSHT [2], LSST [3] and SPLTT [4]) for OPE, SRE and TRE
on the benchmark dataset of [1] in Experiment I (best viewed in color). The numbers in the square
brackets are the ranking scores of the trackers, averaged over all 50 test sequences. Note that the
line style of a curve is determined by the ranking of the corresponding tracker in the plot, not by
the name of the tracker. SPLTT is only evaluated for OPE due to limited computational resources.
For the other 29 compared trackers, readers are referred to [1].
Supplementary Materials
MEEM
Struck
OPE
0.8
ASLA
SCM
3
TLD
TRE
SRE
0.6
0.4
0.2
0 I
V ( OP SV OC DE MB FM IPR OV BC LR
25 R ( (28 C ( F (1 (1 (1 (3 (6 (21 (4)
) 39 ) 29 9) 2) 7) 1) )
)
)
)
IV
O S O D M F I
O B L
(25 PR V (2 CC EF ( B (1 M (1 PR (3 V (6 C (2 R (4
) (39 8) (29 19) 2) 7) 1) ) 1) )
)
)
OPE
IV
O S O D M F I
O B L
(25 PR V (2 CC EF ( B (1 M (1 PR (3 V (6 C (2 R (4
) (39 8) (29 19) 2) 7) 1) ) 1) )
)
)
TRE
SRE
0.6
0.4
0.2
0 IV O S O D M F IP O B L
I
O S O D M F I
O B L
O S O D M F I
I
O B L
(2 PR V (2 CC EF B ( M (1 R ( V (6 C (2 R (4 V (2 PR V (2 CC EF B ( M ( PR ( V ( C ( R (4 V (2 PR V ( CC EF B ( M ( PR V ( C ( R (4
5) (3 8 (2 (1 12 7 31 )
5) (3 8 (2 (1 12 17 31 6) 21 )
5) (3 28 (2 (1 12 17 (31 6) 21 )
1) )
)
)
)
9 )
)
)
9)
9)
)
)
9)
9) 9)
)
)
)
9)
9) 9)
)
)
)
)
Fig. 2. Average precision plot ranking scores (top) and AUC ranking scores (bottome) of the five
leading trackers on different subsets of test sequences in OPE, TRE and SRE on the benchmark
dataset of [1] in Experiment I (best viewed in color). Each subset of sequences corresponds to an
attribute, such as illumination variation (IV), out-of-plane rotation (OPR), scale variation (SV),
occlusion (OCC), deformation (DEF), motion blur (MB), fast motion (FM), in-plane rotation
(IPR), out-of-view (OV), background clutter (BC), low resolution (LR). The number after each
attribute name is the number of sequences that have this attribute. Trackers displayed here are
selected based on their AUC ranking scores in SRE.
TRE
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
0.23
0.02
1.00
0.90
1.00
0.64
0.82
0.77
0.73
0.76
1.00
0.63
1.00
0.96
0.68
0.50
0.48
0.94
0.64
0.50
0.67
0.75
0.81
0.66
0.43
0.99
0.07
0.54
0.24
1.00
0.63
0.41
0.17
1.00
0.17
0.93
0.25
0.66
0.03
0.55
0.06
0.16
0.85
0.63
0.98
0.51
0.49
0.79
1.00
0.79
0.99
0.64
0.42
0.06
0.43
0.87
0.96
0.73
0.18
0.30
1.00
0.98
1.00
0.53
0.04
0.98
0.90
0.66
0.33
0.87
0.95
0.47
0.78
0.84
0.73
0.96
0.35
0.87
0.17
0.23
0.24
0.18
0.17
0.26
0.18
1.00
0.08
0.97
0.42
0.97
0.47
0.78
0.13
0.22
0.67
0.73
0.84
0.16
0.14
0.90
1.00
0.75
0.57
0.58
SRE
0.08
0.08
0.46
0.66
0.98
0.72
0.32
0.11
1.00
0.70
0.99
0.47
0.09
0.91
0.96
0.71
0.75
0.74
0.84
0.46
0.87
0.74
0.71
0.96
0.47
0.81
0.13
0.29
0.44
0.20
0.17
0.33
0.20
1.00
0.05
0.92
0.61
0.97
0.13
0.73
0.13
0.22
0.68
0.82
0.83
0.14
0.16
0.89
0.95
0.95
0.76
0.57
D
0.80
0.27
1.00
0.56
0.90
0.68
0.81
0.97
0.89
0.97
1.00
0.96
0.99
0.96
0.98
0.59
0.53
0.91
0.90
0.48
0.99
1.00
0.92
0.95
0.79
0.94
0.46
0.65
0.89
1.00
0.82
0.77
0.48
0.99
0.04
0.92
0.58
0.31
0.26
0.68
0.88
0.34
0.94
0.76
0.94
0.83
0.57
0.96
1.00
0.44
0.93
0.77
TL
LA
SC
M
AS
M
0.97
0.90
0.96
0.68
0.72
0.52
0.40
0.22
0.63
0.92
1.00
0.44
0.13
0.83
0.96
0.79
0.69
0.84
1.00
0.45
0.67
0.99
0.78
0.84
0.45
0.75
0.23
0.62
0.76
0.39
0.60
0.26
0.20
0.99
0.14
0.92
0.88
0.93
0.83
0.64
0.36
0.52
0.45
0.38
0.74
0.14
0.37
0.57
1.00
0.78
0.35
0.64
ck
0.98
0.83
0.96
0.76
0.62
0.51
0.37
0.41
0.63
0.89
1.00
0.51
0.14
0.80
0.95
0.77
0.70
0.86
0.91
0.45
0.67
0.99
0.75
0.85
0.44
0.79
0.24
0.62
0.76
0.42
0.57
0.28
0.22
0.99
0.15
0.95
0.89
0.96
0.89
0.58
0.47
0.53
0.46
0.39
0.75
0.19
0.32
0.58
1.00
0.68
0.40
0.64
ru
0.46
0.45
0.36
1.00
1.00
0.58
0.28
0.35
1.00
0.94
0.99
0.69
0.09
0.89
0.88
0.59
0.99
0.80
0.99
0.37
0.81
0.85
0.73
0.95
0.56
0.93
0.15
0.70
1.00
0.22
0.18
0.30
0.22
1.00
0.16
0.81
0.77
1.00
0.57
0.78
0.23
0.28
0.93
0.86
0.83
0.14
0.14
0.80
1.00
1.00
0.73
0.65
M
EE
0.46
0.36
1.00
1.00
1.00
0.52
0.65
0.59
0.94
0.74
1.00
0.67
1.00
0.96
0.81
0.73
0.90
0.99
0.97
0.42
0.62
0.86
0.77
0.75
0.33
0.99
0.13
0.78
0.76
1.00
0.52
0.41
0.19
1.00
0.38
0.91
0.55
0.89
0.60
0.45
0.19
0.14
0.57
0.53
0.99
0.60
0.59
0.86
1.00
0.99
1.00
0.71
St
VT
0.90
1.00
1.00
0.90
0.86
0.60
0.82
1.00
0.94
0.99
1.00
0.89
1.00
0.95
0.98
0.57
0.79
0.84
1.00
0.46
0.92
1.00
0.81
0.98
0.83
0.92
0.38
0.70
0.91
1.00
0.86
0.92
0.55
0.99
0.13
0.86
0.97
0.54
0.76
0.64
1.00
0.41
1.00
0.60
0.89
0.81
0.73
0.99
1.00
0.84
1.00
0.83
S
VT
D
OPE
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
SC
M
M
0.03
0.31
1.00
0.87
0.64
0.85
0.68
1.00
0.62
1.00
1.00
0.11
0.73
1.00
0.98
0.60
0.20
0.86
1.00
0.51
0.80
0.55
0.54
0.77
0.41
0.92
0.12
0.97
0.86
1.00
0.86
0.59
0.16
0.98
0.12
0.26
0.41
1.00
0.07
0.32
0.12
0.11
0.25
0.91
0.95
0.46
0.39
0.53
0.96
0.43
0.19
0.61
ck
0.66
0.03
0.44
0.97
1.00
0.65
0.43
0.11
1.00
1.00
1.00
0.50
0.03
0.98
0.98
0.88
0.93
0.86
0.86
0.53
0.77
0.57
0.98
1.00
0.51
1.00
0.16
0.23
1.00
0.15
0.17
0.28
0.35
1.00
0.04
0.97
0.81
1.00
0.11
0.77
0.14
0.27
1.00
0.98
0.95
0.13
0.11
0.87
1.00
1.00
0.94
0.65
M
EE
0.12
0.02
1.00
0.99
1.00
0.65
0.95
0.74
1.00
0.33
1.00
0.34
1.00
1.00
0.92
0.90
0.58
1.00
1.00
0.64
0.75
1.00
0.80
0.79
0.37
1.00
0.11
0.24
0.25
1.00
0.63
0.39
0.12
1.00
0.09
0.92
0.19
0.64
0.04
0.47
0.04
0.25
0.98
0.57
0.99
0.17
0.63
0.88
1.00
0.98
1.00
0.66
St
ru
0.42
0.02
1.00
0.61
0.99
0.86
0.93
1.00
1.00
0.83
1.00
0.89
1.00
0.99
0.99
0.85
0.75
0.85
1.00
0.64
0.30
0.70
0.64
0.82
0.30
0.79
0.08
0.93
0.98
1.00
0.88
0.79
0.26
1.00
0.09
0.94
0.15
0.32
0.69
0.54
0.07
0.14
1.00
0.97
0.91
0.97
0.91
0.95
1.00
0.39
0.83
0.72
D
TL
ck
SC
M
St
1.00
0.97
1.00
0.83
1.00
0.64
0.95
1.00
1.00
0.99
1.00
0.99
1.00
0.99
0.98
0.80
0.65
0.87
1.00
0.60
0.99
1.00
0.93
0.98
0.96
1.00
0.55
0.96
0.97
1.00
0.93
0.97
0.65
1.00
0.05
0.91
0.96
0.35
0.04
0.69
1.00
0.40
1.00
0.63
0.95
0.81
0.50
0.96
1.00
0.37
1.00
0.84
ru
SP
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
M
EE
M
Supplementary Materials
LT
T
4
0.11
0.21
1.00
0.53
0.66
0.65
0.54
0.89
0.67
0.90
0.84
0.17
0.73
0.99
0.97
0.46
0.20
0.71
0.93
0.50
0.79
0.58
0.70
0.84
0.42
0.95
0.10
0.88
0.87
1.00
0.37
0.61
0.16
0.98
0.12
0.33
0.37
0.58
0.08
0.44
0.11
0.12
0.56
0.91
0.88
0.33
0.33
0.53
0.60
0.50
0.52
0.57
Fig. 3. Precision plot ranking scores of the top five trackers in OPE, TRE and SRE on the benchmark dataset of [1] in Experiment I. The trackers shown in this figure are SPLTT [4], Struck [5],
SCM [6], TLD [7], VTD [8], VTS [9] and ASLA [10]. Our tracker tends to better handle the
sequences like “basketball”, “bolt”, “david3”, “football”, “football1”, “freeman4”, “ironman”,
“lemming”, “liquor”, “matrix”, “skiing”, “soccer”, “subway”, “tiger1” and “tiger2”, in which the
target often undergoes different levels of occlusion, non-rigid motion and out-of-plane rotation.
OPE
0.21
0.01
0.75
0.49
0.87
0.42
0.66
0.53
0.67
0.25
0.85
0.29
0.73
0.54
0.54
0.72
0.72
0.77
0.84
0.60
0.53
0.66
0.34
0.26
0.17
0.73
0.09
0.17
0.20
0.61
0.48
0.40
0.10
0.80
0.16
0.70
0.36
0.36
0.04
0.31
0.04
0.19
0.65
0.51
0.71
0.15
0.54
0.61
0.57
0.51
0.72
0.47
0.58
0.74
0.70
0.60
0.67
0.37
0.63
0.64
0.65
0.64
0.80
0.63
0.75
0.50
0.66
0.61
0.75
0.66
0.81
0.55
0.61
0.69
0.43
0.37
0.43
0.55
0.26
0.50
0.68
0.69
0.64
0.74
0.35
0.72
0.14
0.64
0.67
0.38
0.60
0.39
0.40
0.28
0.75
0.49
0.64
0.64
0.63
0.74
0.62
0.53
0.67
0.59
0.36
0.27
0.67
0.65
0.84
0.37
0.55
0.48
0.66
0.52
0.82
0.49
0.77
0.50
0.57
0.64
0.77
0.73
0.83
0.51
0.41
0.55
0.40
0.32
0.19
0.65
0.12
0.55
0.62
0.68
0.40
0.38
0.15
0.76
0.33
0.67
0.44
0.46
0.50
0.29
0.10
0.11
0.43
0.45
0.69
0.50
0.51
0.66
0.61
0.60
0.70
0.51
0.34
0.34
0.28
0.82
0.83
0.55
0.27
0.25
0.78
0.70
0.77
0.51
0.11
0.63
0.73
0.62
0.84
0.66
0.83
0.53
0.48
0.51
0.43
0.61
0.37
0.65
0.12
0.55
0.83
0.17
0.16
0.31
0.17
0.82
0.17
0.59
0.60
0.83
0.52
0.51
0.13
0.21
0.71
0.69
0.59
0.19
0.17
0.67
0.75
0.81
0.53
0.51
TRE
0.36
0.34
0.25
0.81
0.79
0.54
0.23
0.21
0.77
0.67
0.82
0.31
0.14
0.64
0.80
0.63
0.52
0.61
0.84
0.52
0.45
0.54
0.42
0.65
0.32
0.61
0.11
0.51
0.63
0.19
0.20
0.32
0.14
0.84
0.21
0.69
0.68
0.71
0.44
0.43
0.12
0.22
0.38
0.50
0.56
0.22
0.19
0.70
0.77
0.66
0.51
0.48
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging-1
jogging-2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
0.55
0.16
0.69
0.33
0.67
0.39
0.58
0.58
0.57
0.51
0.75
0.68
0.67
0.53
0.56
0.65
0.67
0.66
0.67
0.56
0.64
0.67
0.37
0.34
0.36
0.61
0.30
0.46
0.61
0.65
0.59
0.68
0.31
0.70
0.10
0.62
0.49
0.29
0.20
0.38
0.36
0.24
0.60
0.56
0.64
0.64
0.55
0.57
0.51
0.35
0.61
0.52
0.18
0.01
0.67
0.44
0.73
0.40
0.57
0.52
0.46
0.43
0.74
0.46
0.67
0.53
0.42
0.61
0.67
0.67
0.54
0.55
0.43
0.46
0.35
0.23
0.20
0.67
0.07
0.37
0.17
0.64
0.47
0.43
0.11
0.73
0.25
0.64
0.27
0.33
0.04
0.32
0.04
0.14
0.55
0.49
0.64
0.46
0.45
0.50
0.53
0.46
0.67
0.44
0.29
0.03
0.32
0.62
0.68
0.58
0.17
0.21
0.74
0.65
0.67
0.38
0.05
0.67
0.69
0.65
0.47
0.64
0.71
0.55
0.49
0.47
0.42
0.63
0.17
0.53
0.13
0.16
0.16
0.11
0.13
0.31
0.16
0.76
0.13
0.67
0.39
0.66
0.35
0.49
0.08
0.16
0.44
0.53
0.56
0.20
0.15
0.67
0.70
0.54
0.35
0.42
0.05
0.06
0.34
0.49
0.72
0.55
0.26
0.09
0.71
0.49
0.63
0.35
0.10
0.62
0.71
0.68
0.70
0.58
0.66
0.54
0.53
0.43
0.43
0.59
0.25
0.52
0.10
0.19
0.29
0.13
0.14
0.33
0.15
0.74
0.10
0.61
0.50
0.75
0.17
0.46
0.08
0.20
0.44
0.58
0.56
0.18
0.18
0.65
0.62
0.70
0.48
0.42
TL
D
5
SC
M
k
AS
LA
M
0.19
0.08
0.64
0.67
0.50
0.54
0.41
0.43
0.45
0.68
0.83
0.23
0.63
0.76
0.73
0.64
0.64
0.61
0.80
0.49
0.44
0.67
0.42
0.68
0.23
0.54
0.07
0.66
0.68
0.61
0.35
0.28
0.08
0.81
0.16
0.42
0.48
0.64
0.10
0.28
0.10
0.14
0.31
0.70
0.48
0.34
0.38
0.57
0.43
0.40
0.21
0.46
St
ru
c
M
EE
CX
T
AS
LA
M
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging-1
jogging-2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
SC
M
EE
M
0.02
0.16
0.65
0.63
0.44
0.45
0.40
0.76
0.40
0.71
0.68
0.10
0.59
0.58
0.57
0.64
0.58
0.61
0.80
0.49
0.49
0.37
0.28
0.44
0.22
0.57
0.11
0.76
0.65
0.66
0.53
0.51
0.16
0.63
0.23
0.20
0.39
0.71
0.22
0.19
0.07
0.13
0.18
0.68
0.67
0.38
0.27
0.48
0.45
0.31
0.13
0.44
St
ru
ck
k
0.46
0.02
0.37
0.75
0.83
0.59
0.32
0.10
0.77
0.71
0.73
0.39
0.07
0.69
0.81
0.76
0.78
0.72
0.74
0.60
0.48
0.40
0.61
0.71
0.26
0.67
0.12
0.18
0.72
0.12
0.14
0.32
0.26
0.79
0.11
0.66
0.68
0.85
0.17
0.47
0.08
0.24
0.71
0.74
0.68
0.16
0.09
0.67
0.70
0.80
0.65
0.50
St
ru
c
0.29
0.01
0.68
0.38
0.73
0.65
0.57
0.66
0.68
0.47
0.72
0.70
0.68
0.73
0.77
0.74
0.72
0.63
0.79
0.55
0.19
0.49
0.41
0.57
0.14
0.56
0.07
0.63
0.71
0.74
0.70
0.64
0.15
0.83
0.27
0.64
0.13
0.27
0.57
0.37
0.04
0.13
0.57
0.72
0.62
0.72
0.66
0.54
0.49
0.29
0.60
0.52
TL
D
0.80
0.43
0.78
0.46
0.84
0.39
0.66
0.60
0.71
0.53
0.83
0.67
0.73
0.55
0.57
0.71
0.72
0.69
0.75
0.60
0.68
0.67
0.38
0.35
0.42
0.70
0.37
0.66
0.63
0.69
0.67
0.77
0.39
0.77
0.10
0.59
0.69
0.29
0.04
0.40
0.41
0.32
0.67
0.53
0.67
0.64
0.54
0.61
0.54
0.27
0.70
0.57
SC
M
M
SP
EE
M
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging-1
jogging-2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
LT
T
Supplementary Materials
0.07
0.12
0.60
0.36
0.39
0.44
0.34
0.61
0.37
0.54
0.55
0.12
0.52
0.57
0.54
0.56
0.53
0.56
0.59
0.47
0.52
0.37
0.33
0.44
0.20
0.55
0.09
0.62
0.60
0.65
0.26
0.58
0.17
0.64
0.20
0.23
0.35
0.63
0.11
0.26
0.06
0.12
0.36
0.67
0.57
0.33
0.28
0.37
0.33
0.41
0.33
0.40
SRE
Fig. 4. AUC ranking scores of the top five trackers in OPE, TRE and SRE on the benchmark
dataset of [1] in Experiment I. The trackers shown in this figure are SPLTT [4], Struck [5],
SCM [6], TLD [7], ASLA [10] and CXT [11]. Our tracker tends to better handle the sequences
like “basketball”, “bolt”, “david3”, “football”, “football1”, “freeman4”, “ironman”, “lemming”,
“liquor”, “matrix”, “skiing”, “soccer”, “subway”, “tiger1” and “tiger2”, in which the target often
undergoes different levels of occlusion, non-rigid motion and out-of-plane rotation.
OPE
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
0.90
1.00
1.00
0.90
0.86
0.60
0.82
1.00
0.94
0.99
1.00
0.89
1.00
0.95
0.98
0.57
0.79
0.84
1.00
0.46
0.92
1.00
0.81
0.98
0.83
0.92
0.38
0.70
0.91
1.00
0.86
0.92
0.55
0.99
0.13
0.86
0.97
0.54
0.76
0.64
1.00
0.41
1.00
0.60
0.89
0.81
0.73
0.99
1.00
0.84
1.00
0.83
0.94
1.00
1.00
0.83
0.86
0.61
0.72
1.00
0.94
0.99
1.00
0.78
1.00
0.95
0.98
0.60
0.79
0.92
1.00
0.45
0.90
1.00
0.87
0.96
0.74
0.92
0.37
0.62
0.83
1.00
0.54
0.90
0.55
0.99
0.13
0.86
0.98
0.52
0.72
0.79
1.00
0.44
1.00
0.64
0.94
0.69
0.73
0.99
1.00
0.80
1.00
0.82
0.91
0.99
1.00
0.88
0.86
0.61
0.74
1.00
0.94
0.99
1.00
0.76
1.00
0.95
0.98
0.60
0.79
0.93
1.00
0.45
0.92
1.00
0.83
0.98
0.86
0.93
0.39
0.62
0.83
1.00
0.54
0.87
0.55
0.99
0.13
0.86
0.98
0.51
0.71
0.72
0.94
0.42
1.00
0.56
0.94
0.69
0.73
0.99
1.00
0.75
1.00
0.82
TRE
0.91
0.99
1.00
0.88
0.86
0.61
0.74
1.00
0.94
0.99
1.00
0.76
1.00
0.95
0.98
0.60
0.79
0.93
1.00
0.45
0.92
1.00
0.83
0.98
0.86
0.93
0.39
0.62
0.83
1.00
0.54
0.87
0.55
0.99
0.13
0.86
0.98
0.51
0.71
0.72
0.94
0.42
1.00
0.56
0.94
0.69
0.73
0.99
1.00
0.75
1.00
0.82
0.53
0.60
1.00
0.90
0.75
0.59
0.85
0.97
1.00
0.98
1.00
0.72
1.00
0.95
0.95
0.57
0.80
0.84
1.00
0.46
0.92
0.98
0.81
0.98
0.83
0.95
0.21
0.81
0.99
1.00
0.70
0.53
0.21
0.99
0.11
0.83
0.96
0.61
0.69
0.66
0.24
0.29
1.00
0.60
0.89
0.70
0.72
0.87
1.00
0.71
0.95
0.77
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
M
M
EE
M
−l
kh
SV
M
−a
vg
SV
M
−b
as
M
e
EE
M
−g
ra
y
0.12
0.05
1.00
0.83
1.00
0.66
0.94
1.00
1.00
0.98
1.00
0.97
1.00
0.99
0.97
0.80
0.69
0.87
1.00
0.60
0.99
1.00
0.93
0.98
0.96
1.00
0.07
0.23
0.93
1.00
0.83
0.71
0.10
1.00
0.04
0.93
0.95
0.34
0.03
0.69
0.14
0.43
1.00
0.63
0.95
0.81
0.52
0.95
1.00
0.43
1.00
0.75
M
EE
0.98
0.90
1.00
0.57
1.00
0.64
0.88
1.00
1.00
0.99
1.00
0.75
1.00
0.99
0.98
0.78
0.65
0.96
1.00
0.60
1.00
1.00
0.41
1.00
0.99
1.00
0.67
0.23
0.97
1.00
0.57
0.97
0.65
1.00
0.05
0.91
0.97
0.35
0.04
0.69
1.00
0.40
1.00
0.79
0.97
0.81
0.50
0.96
1.00
0.39
1.00
0.80
M
0.98
0.90
1.00
0.57
1.00
0.64
0.88
1.00
1.00
0.99
1.00
0.75
1.00
0.99
0.98
0.78
0.65
0.96
1.00
0.60
1.00
1.00
0.41
1.00
0.99
1.00
0.67
0.23
0.97
1.00
0.57
0.97
0.65
1.00
0.05
0.91
0.97
0.35
0.04
0.69
1.00
0.40
1.00
0.79
0.97
0.81
0.50
0.96
1.00
0.39
1.00
0.80
M
−l
kh
SV
M
−a
vg
SV
M
−b
as
M
e
EE
M
−g
ra
y
M
EE
1.00
0.98
1.00
0.71
1.00
0.71
0.87
1.00
1.00
0.99
1.00
0.99
1.00
0.99
0.99
0.80
0.65
0.87
1.00
0.60
0.99
1.00
0.93
0.97
0.96
0.92
0.65
0.23
0.97
1.00
0.26
0.89
0.66
1.00
0.06
0.91
0.99
0.36
0.04
0.88
1.00
0.27
1.00
0.79
0.95
0.82
0.50
0.96
1.00
0.45
1.00
0.81
M
EE
SV
1.00
0.97
1.00
0.83
1.00
0.64
0.95
1.00
1.00
0.99
1.00
0.99
1.00
0.99
0.98
0.80
0.65
0.87
1.00
0.60
0.99
1.00
0.93
0.98
0.96
1.00
0.55
0.96
0.97
1.00
0.93
0.97
0.65
1.00
0.05
0.91
0.96
0.35
0.04
0.69
1.00
0.40
1.00
0.63
0.95
0.81
0.50
0.96
1.00
0.37
1.00
0.84
M
EE
as
e
M
−g
ra
y
vg
SV
M
−b
M
EE
M
−a
M
EE
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
M
−l
kh
Supplementary Materials
M
6
0.80
0.27
1.00
0.56
0.90
0.68
0.81
0.97
0.89
0.97
1.00
0.96
0.99
0.96
0.98
0.59
0.53
0.91
0.90
0.48
0.99
1.00
0.92
0.95
0.79
0.94
0.46
0.65
0.89
1.00
0.82
0.77
0.48
0.99
0.04
0.92
0.58
0.31
0.26
0.68
0.88
0.34
0.94
0.76
0.94
0.83
0.57
0.96
1.00
0.44
0.93
0.77
0.78
0.27
1.00
0.41
0.98
0.68
0.78
0.97
0.89
0.97
1.00
0.85
0.99
0.97
0.98
0.59
0.53
0.92
0.98
0.49
1.00
1.00
0.98
0.97
0.86
0.95
0.46
0.35
0.48
1.00
0.47
0.77
0.47
0.99
0.04
0.93
0.64
0.29
0.25
0.71
0.88
0.37
0.88
0.69
0.95
0.83
0.59
0.97
0.94
0.41
1.00
0.75
0.80
0.27
1.00
0.45
0.98
0.68
0.78
0.97
0.89
0.97
1.00
0.79
0.99
0.96
0.98
0.61
0.53
0.95
0.98
0.48
0.97
1.00
0.96
0.97
0.85
0.96
0.45
0.35
0.49
1.00
0.50
0.79
0.50
0.99
0.05
0.93
0.54
0.29
0.26
0.71
0.88
0.28
0.88
0.65
0.94
0.83
0.59
0.96
0.94
0.48
1.00
0.75
0.77
0.27
1.00
0.45
0.93
0.68
0.79
0.97
0.89
0.97
1.00
0.83
0.99
0.96
0.98
0.60
0.53
0.93
0.98
0.49
0.98
1.00
0.97
0.97
0.83
0.96
0.45
0.35
0.48
1.00
0.52
0.78
0.48
0.99
0.05
0.92
0.54
0.30
0.26
0.75
0.88
0.30
0.88
0.70
0.95
0.83
0.59
0.96
1.00
0.49
0.93
0.75
0.43
0.13
1.00
0.56
0.95
0.69
0.94
0.94
1.00
0.94
1.00
0.83
1.00
0.96
0.96
0.59
0.50
0.91
0.90
0.48
0.99
0.99
0.92
0.95
0.79
0.97
0.10
0.83
0.96
1.00
0.77
0.52
0.14
0.99
0.04
0.92
0.48
0.34
0.23
0.54
0.12
0.23
1.00
0.76
0.94
0.77
0.56
0.94
1.00
0.36
1.00
0.72
SRE
Fig. 5. Precision plot ranking scores of MEEM and its baselines in OPE, TRE and SRE on the
benchmark dataset of [1] in Experiment I. Besides the baselines evaluated in our paper, we also
report another baseline MEEM-gray, which is the full implementation of MEEM, but only uses the gray-scale images as input. MEEM-gray generally gives lower scores than MEEM, but
on average it still compares favorably with the state-of-the-art trackers in Fig. 3. Significant improvement of MEEM over SVM-base is observed on such sequences as “david3” (OPE, TRE and
SRE), “lemming” (OPE, TRE and SRE), “jogging-1” (OPE and SRE) and “jogging2” (SRE),
where challenges like occlusions and large appearance variations, could lead to model drift.
0.80
0.43
0.78
0.46
0.84
0.39
0.66
0.60
0.71
0.53
0.83
0.67
0.73
0.55
0.57
0.71
0.72
0.69
0.75
0.60
0.68
0.67
0.38
0.35
0.42
0.70
0.37
0.66
0.63
0.69
0.67
0.77
0.39
0.77
0.10
0.59
0.69
0.29
0.04
0.40
0.41
0.32
0.67
0.53
0.67
0.64
0.54
0.61
0.54
0.27
0.70
0.57
0.69
0.46
0.78
0.47
0.84
0.42
0.61
0.60
0.70
0.53
0.84
0.69
0.73
0.55
0.57
0.71
0.72
0.69
0.75
0.64
0.69
0.67
0.38
0.35
0.41
0.64
0.45
0.17
0.62
0.69
0.22
0.73
0.41
0.77
0.11
0.59
0.69
0.31
0.04
0.46
0.41
0.26
0.67
0.64
0.67
0.64
0.53
0.61
0.54
0.45
0.71
0.56
0.70
0.43
0.78
0.47
0.84
0.39
0.63
0.60
0.71
0.53
0.83
0.53
0.73
0.55
0.57
0.71
0.72
0.72
0.75
0.64
0.72
0.67
0.21
0.35
0.43
0.70
0.45
0.17
0.63
0.69
0.42
0.76
0.40
0.77
0.09
0.59
0.69
0.30
0.04
0.39
0.41
0.29
0.67
0.64
0.69
0.64
0.54
0.61
0.54
0.45
0.71
0.56
OPE
0.70
0.43
0.78
0.47
0.84
0.39
0.63
0.60
0.71
0.53
0.83
0.53
0.73
0.55
0.57
0.71
0.72
0.72
0.75
0.64
0.72
0.67
0.21
0.35
0.43
0.70
0.45
0.17
0.63
0.69
0.42
0.76
0.40
0.77
0.09
0.59
0.69
0.30
0.04
0.39
0.41
0.29
0.67
0.64
0.69
0.64
0.54
0.61
0.54
0.45
0.71
0.56
0.08
0.03
0.78
0.46
0.85
0.39
0.68
0.62
0.71
0.53
0.83
0.67
0.73
0.55
0.56
0.71
0.74
0.69
0.75
0.60
0.68
0.81
0.38
0.35
0.42
0.70
0.07
0.16
0.60
0.69
0.61
0.61
0.08
0.77
0.09
0.59
0.70
0.34
0.04
0.39
0.09
0.32
0.56
0.53
0.67
0.64
0.49
0.61
0.54
0.45
0.73
0.52
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
M
EE
M
M
EE
M
−l
kh
SV
M
−a
vg
SV
M
−b
as
M
e
EE
M
−g
ra
y
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
7
M
EE
M
M
EE
M
−l
kh
SV
M
−a
vg
SV
M
−b
as
M
e
EE
M
−g
ra
y
EE
M
−l
kh
SV
M
−a
vg
SV
M
−b
as
M
e
EE
M
−g
ra
y
M
M
EE
M
Supplementary Materials
0.58
0.74
0.70
0.60
0.67
0.37
0.63
0.64
0.65
0.64
0.80
0.63
0.75
0.50
0.66
0.61
0.75
0.66
0.81
0.55
0.61
0.69
0.43
0.37
0.43
0.55
0.26
0.50
0.68
0.69
0.64
0.74
0.35
0.72
0.14
0.64
0.67
0.38
0.60
0.39
0.40
0.28
0.75
0.49
0.64
0.64
0.63
0.74
0.62
0.53
0.67
0.59
0.55
0.16
0.69
0.33
0.67
0.39
0.58
0.58
0.57
0.51
0.75
0.68
0.67
0.53
0.56
0.65
0.67
0.66
0.67
0.56
0.64
0.67
0.37
0.34
0.36
0.61
0.30
0.46
0.61
0.65
0.59
0.68
0.31
0.70
0.10
0.62
0.49
0.29
0.20
0.38
0.36
0.24
0.60
0.56
0.64
0.64
0.55
0.57
0.51
0.35
0.61
0.52
0.61
0.75
0.70
0.58
0.67
0.37
0.56
0.64
0.65
0.64
0.81
0.55
0.75
0.50
0.66
0.63
0.75
0.70
0.81
0.57
0.60
0.69
0.45
0.37
0.40
0.55
0.26
0.45
0.62
0.69
0.42
0.73
0.34
0.72
0.14
0.64
0.67
0.38
0.57
0.46
0.40
0.29
0.75
0.51
0.67
0.57
0.63
0.74
0.62
0.54
0.66
0.58
0.57
0.74
0.70
0.60
0.67
0.37
0.58
0.64
0.65
0.64
0.80
0.54
0.75
0.50
0.66
0.63
0.75
0.70
0.81
0.57
0.61
0.69
0.43
0.37
0.43
0.56
0.27
0.45
0.62
0.69
0.43
0.70
0.34
0.72
0.14
0.64
0.67
0.38
0.57
0.42
0.38
0.28
0.75
0.46
0.67
0.56
0.63
0.74
0.62
0.52
0.65
0.57
TRE
0.57
0.74
0.70
0.60
0.67
0.37
0.58
0.64
0.65
0.64
0.80
0.54
0.75
0.50
0.66
0.63
0.75
0.70
0.81
0.57
0.61
0.69
0.43
0.37
0.43
0.56
0.27
0.45
0.62
0.69
0.43
0.70
0.34
0.72
0.14
0.64
0.67
0.38
0.57
0.42
0.38
0.28
0.75
0.46
0.67
0.56
0.63
0.74
0.62
0.52
0.65
0.57
0.37
0.45
0.70
0.60
0.57
0.37
0.67
0.64
0.70
0.64
0.80
0.50
0.75
0.50
0.65
0.61
0.75
0.66
0.81
0.55
0.61
0.69
0.43
0.37
0.43
0.59
0.16
0.57
0.74
0.69
0.54
0.48
0.16
0.72
0.13
0.63
0.67
0.41
0.55
0.41
0.13
0.20
0.71
0.49
0.64
0.57
0.63
0.66
0.63
0.49
0.65
0.55
basketball
bolt
boy
car4
carDark
carScale
coke
couple
crossing
david
david2
david3
deer
dog1
doll
dudek
faceocc1
faceocc2
fish
fleetface
football
football1
freeman1
freeman3
freeman4
girl
ironman
jogging−1
jogging−2
jumping
lemming
liquor
matrix
mhyang
motorRolling
mountainBike
shaking
singer1
singer2
skating1
skiing
soccer
subway
suv
sylvester
tiger1
tiger2
trellis
walking
walking2
woman
AVG
0.54
0.16
0.69
0.31
0.72
0.39
0.56
0.58
0.57
0.50
0.75
0.61
0.67
0.53
0.56
0.65
0.67
0.66
0.70
0.59
0.65
0.67
0.39
0.34
0.39
0.60
0.29
0.24
0.31
0.65
0.35
0.68
0.29
0.70
0.10
0.62
0.46
0.29
0.20
0.40
0.36
0.25
0.57
0.51
0.65
0.64
0.55
0.58
0.48
0.40
0.66
0.50
0.56
0.16
0.69
0.30
0.72
0.39
0.56
0.58
0.57
0.51
0.75
0.58
0.67
0.53
0.57
0.66
0.67
0.67
0.70
0.57
0.62
0.67
0.38
0.34
0.38
0.61
0.29
0.24
0.32
0.65
0.38
0.69
0.31
0.70
0.10
0.62
0.47
0.29
0.20
0.40
0.36
0.22
0.57
0.49
0.63
0.64
0.55
0.57
0.48
0.43
0.66
0.50
SRE
0.53
0.16
0.69
0.30
0.68
0.39
0.56
0.58
0.57
0.51
0.75
0.59
0.67
0.53
0.56
0.66
0.67
0.66
0.70
0.58
0.64
0.67
0.38
0.34
0.38
0.61
0.29
0.24
0.34
0.65
0.39
0.69
0.31
0.70
0.10
0.62
0.47
0.29
0.20
0.40
0.36
0.22
0.57
0.51
0.64
0.64
0.55
0.57
0.51
0.43
0.61
0.50
0.37
0.08
0.69
0.33
0.70
0.39
0.65
0.58
0.65
0.49
0.75
0.59
0.66
0.53
0.56
0.65
0.67
0.66
0.67
0.56
0.64
0.64
0.37
0.34
0.36
0.64
0.08
0.56
0.62
0.65
0.56
0.46
0.11
0.70
0.11
0.61
0.43
0.30
0.18
0.31
0.07
0.17
0.60
0.56
0.64
0.61
0.52
0.57
0.50
0.34
0.65
0.49
Fig. 6. AUC ranking scores of MEEM and its baselines in OPE, TRE and SRE on the benchmark
dataset of [1] in Experiment I. Besides the baselines evaluated in our paper, we also report another
baseline MEEM-gray, which is the full implementation of MEEM, but only uses the gray-scale
images as input. MEEM-gray generally gives lower scores than MEEM, but on average it still
compares favorably with most of the top trackers in Fig. 4. Significant improvement of MEEM
over SVM-base is observed on such sequences as “david3” (OPE, TRE and SRE), “lemming”
(OPE, TRE and SRE), “jogging-1” (OPE and SRE) and “jogging2” (SRE), where challenges like
occlusions and large appearance variations, could lead to model drift.
ro
ck
y
bi
lli
AV
0.74
0.71
0.59
0.59
0.22
0.56
0.47
1.00
0.59
0.59
0.59
0.56
0.53
0.68
0.96
0.15
0.15
0.15
0.95
0.23
0.04
0.99
0.99
0.99
0.99
1.00
0.98
0.32
0.78
0.79
0.89
0.89
0.15
0.12
0.53
0.90
0.74
0.70
0.64
0.49
0.41
0.29
OPE
0.90
0.94
0.94
0.94
0.10
0.38
0.14
0.58
0.65
0.64
0.64
0.59
0.39
0.17
0.56
0.37
0.30
0.30
0.19
0.15
0.17
0.75
0.53
0.49
0.49
0.43
0.34
0.07
0.69
0.65
0.57
0.57
0.22
0.39
0.21
0.98
0.70
0.74
0.74
0.79
0.94
0.56
0.76
0.14
0.13
0.13
0.26
0.43
0.05
0.88
0.89
0.96
0.96
1.00
0.70
0.27
0.84
0.83
0.84
0.84
0.24
0.26
0.57
0.77
0.64
0.63
0.63
0.41
0.42
0.25
SVM−avg 0.77
SVM−base 0.77
ASLA 0.21
Struck 0.29
TLD 0.32
SRE
0.74
0.67
0.28
0.28
0.42
0.39
0.36
0.73
0.54
0.63
0.55
0.37
0.30
0.07
0.66
0.62
0.54
0.54
0.24
0.52
0.49
0.81
0.48
0.48
0.48
0.43
0.45
0.42
Je
an
AV
G
0.72
0.76
0.73
0.29
0.67
0.34
0.33
llie
0.78
0.78
0.78
0.78
0.09
0.36
0.13
0.63
0.22
0.22
0.22
0.71
0.57
0.04
0.77
0.79
0.80
0.80
0.78
0.63
0.27
0.56
0.58
0.62
0.62
0.13
0.11
0.41
0.72
0.62
0.58
0.53
0.40
0.40
0.29
0.71 0.54 0.63 0.59 0.63 0.73 0.53 0.61 0.56 0.62
0.74 0.60 0.45 0.43 0.61 0.52 0.18 0.61 0.57 0.54
SVM−avg 0.66 0.73 0.59 0.36 0.40 0.53 0.55 0.16 0.66 0.57 0.52
OPE
MEEM 0.78
MEEM−lkh 0.79
bi
la
tin
TLD 0.27
ro
Struck 0.20
pe
d
ASLA 0.29
g1
bo
xin
g2
pe
d1
SVM−avg 0.71
SVM−base 0.71
xin
MEEM 0.74
MEEM−lkh 0.71
bo
TLD 0.33
ac
e
da
nc
e
Struck 0.27
ca
rR
ASLA 0.34
ba
ll
SVM−avg 0.90
SVM−base 0.90
G
pe
d2
0.94
0.69
0.81
0.71
0.50
0.36
0.06
ck
y
bo
xin
g1
bo
xin
g2
pe
d1
0.83
0.73
0.24
0.24
0.39
0.27
0.18
2
ca
rR
ac
0.82
0.87
0.87
0.34
0.72
0.37
0.16
da
nc
e
ba
ll
0.96
0.96
0.96
0.95
0.07
0.36
0.15
la
tin
MEEM 0.98
MEEM−lkh 0.96
eJ
ea
n
Supplementary Materials
e
8
MEEM 0.65
SVM−base 0.66
0.73
0.13
Struck 0.26 0.34
TLD 0.31 0.13
ASLA 0.28
0.59
0.55
0.39
0.39
0.36
0.35
0.28
0.36
0.40
0.34
0.29
0.10
0.53
0.26
0.43
0.39
0.55
0.59
0.73
0.36
0.16
0.27
0.45
0.06
0.66
0.74
0.48
0.18
0.57
0.19
0.21
0.41
0.52
0.37
0.39
0.27
SRE
MEEM−lkh 0.67
Fig. 7. Precision plot ranking scores (top) and AUC ranking scores (bottom) of MEEM, MEEMlkh, SVM-avg, SVM-base and other state-of-the-art trackers on our newly collected sequences in
Experiment II.