This table lists the benchmark results for the low-res two-view scenario. This benchmark evaluates the Middlebury stereo metrics (for all metrics, smaller is better):

The mask determines whether the metric is evaluated for all pixels with ground truth, or only for pixels which are visible in both images (non-occluded).
The coverage selector allows to limit the table to results for all pixels (dense), or a given minimum fraction of pixels.

Methods with suffix _ROB may participate in the Robust Vision Challenge.

Click one or more dataset result cells or column headers to show visualizations. Most visualizations are only available for training datasets. The visualizations may not work with mobile browsers.




Method Infoalllakes. 1llakes. 1ssand box 1lsand box 1sstora. room 1lstora. room 1sstora. room 2lstora. room 2sstora. room 2 1lstora. room 2 1sstora. room 2 2lstora. room 2 2sstora. room 3lstora. room 3stunnel 1ltunnel 1stunnel 2ltunnel 2stunnel 3ltunnel 3s
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
PMTNettwo views0.21
1
0.19
20
0.85
2
0.02
15
0.02
37
1.14
2
0.30
69
0.88
1
0.11
1
0.03
1
0.02
3
0.07
5
0.02
4
0.37
8
0.12
5
0.00
1
0.00
1
0.00
1
0.00
1
0.03
22
0.00
1
CREStereotwo views0.22
2
0.17
14
0.64
1
0.00
1
0.01
26
1.22
4
0.01
6
1.15
2
0.56
6
0.10
2
0.07
6
0.04
1
0.00
1
0.32
7
0.10
4
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.00
1
Gwc-CoAtRStwo views0.42
3
0.12
9
1.28
10
0.03
24
0.13
68
1.03
1
0.02
13
2.33
4
0.15
4
0.35
5
0.01
1
2.27
44
0.03
6
0.20
2
0.25
6
0.00
1
0.01
74
0.00
1
0.22
121
0.01
7
0.01
13
R-Stereo Traintwo views0.44
4
0.09
5
0.94
4
0.02
15
0.00
1
1.24
6
0.09
40
3.65
39
0.11
1
1.87
44
0.42
30
0.06
3
0.01
2
0.25
3
0.05
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
37
RAFT-Stereopermissivetwo views0.44
4
0.09
5
0.94
4
0.02
15
0.00
1
1.24
6
0.09
40
3.65
39
0.11
1
1.87
44
0.42
30
0.06
3
0.01
2
0.25
3
0.05
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
37
Lahav Lipson, Zachary Teed, and Jia Deng: RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching. 3DV
FENettwo views0.52
6
0.07
1
1.62
24
0.04
30
0.00
1
1.19
3
0.05
32
3.10
18
0.98
15
0.89
21
0.01
1
1.63
26
0.09
8
0.47
10
0.28
7
0.00
1
0.01
74
0.00
1
0.00
1
0.00
1
0.01
13
ACVNettwo views0.57
7
0.21
22
1.34
14
0.09
46
0.00
1
1.36
19
0.22
60
2.40
5
1.05
17
0.63
11
0.66
41
1.94
33
0.26
27
0.31
5
0.79
34
0.00
1
0.00
1
0.00
1
0.00
1
0.08
54
0.00
1
DPM-Stereotwo views0.58
8
0.23
28
0.90
3
0.08
41
0.00
1
1.43
32
0.01
6
6.01
109
1.38
37
0.92
23
0.11
9
0.04
1
0.19
19
0.12
1
0.07
3
0.00
1
0.01
74
0.00
1
0.00
1
0.03
22
0.01
13
AdaStereotwo views0.65
9
0.09
5
1.12
7
0.02
15
0.00
1
1.45
34
0.23
63
4.38
69
1.44
46
1.17
34
0.37
27
1.56
22
0.18
17
0.56
11
0.37
12
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.00
1
Xiao Song, Guorun Yang, Xinge Zhu, Hui Zhou, Zhe Wang, Jianping Shi: AdaStereo: A Simple and Efficient Approach for Adaptive Stereo Matching. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
DMCAtwo views0.68
10
0.22
24
1.96
35
0.01
5
0.01
26
1.28
8
0.02
13
3.02
14
1.54
55
2.14
50
0.15
10
1.74
28
0.10
11
0.45
9
0.95
40
0.00
1
0.00
1
0.00
1
0.00
1
0.02
18
0.01
13
CFNet-ftpermissivetwo views0.77
11
0.46
49
1.34
14
0.01
5
0.27
89
1.32
13
0.02
13
2.40
5
0.68
7
0.63
11
0.18
13
3.23
67
0.21
22
3.97
28
0.54
22
0.00
1
0.00
1
0.01
88
0.00
1
0.03
22
0.02
37
DN-CSS_ROBtwo views0.77
11
0.90
100
2.01
39
0.85
114
0.00
1
1.47
36
0.01
6
2.92
13
0.93
14
0.12
3
0.53
37
0.44
7
0.16
15
4.33
36
0.35
10
0.00
1
0.00
1
0.00
1
0.00
1
0.30
108
0.01
13
CFNet_RVCtwo views0.77
11
0.46
49
1.34
14
0.01
5
0.27
89
1.32
13
0.02
13
2.40
5
0.68
7
0.63
11
0.18
13
3.23
67
0.21
22
3.97
28
0.54
22
0.00
1
0.00
1
0.01
88
0.00
1
0.03
22
0.02
37
acv_fttwo views0.79
14
0.21
22
1.28
10
0.01
5
1.08
120
1.68
52
0.01
6
3.75
43
1.05
17
1.31
36
0.66
41
1.94
33
0.46
42
0.31
5
1.68
56
0.00
1
0.00
1
0.00
1
0.00
1
0.27
101
0.00
1
HITNettwo views0.80
15
0.58
67
2.14
43
0.69
109
0.00
1
1.35
17
0.02
13
4.50
74
1.15
24
1.76
41
0.16
11
0.60
8
0.30
30
1.13
12
1.64
54
0.00
1
0.00
1
0.00
1
0.00
1
0.06
47
0.00
1
Vladimir Tankovich, Christian Häne, Yinda Zhang, Adarsh Kowdle, Sean Fanello, Sofien Bouaziz: HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matching. CVPR 2021
STTStereotwo views0.83
16
0.34
34
2.44
48
0.24
73
0.15
73
1.41
28
0.10
44
3.02
14
1.55
56
1.34
37
0.17
12
1.77
29
0.45
41
3.42
19
0.30
8
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.01
13
StereoDRNet-Refinedtwo views0.83
16
0.12
9
1.18
9
0.02
15
0.01
26
1.31
12
0.00
1
3.33
25
1.07
20
1.92
47
1.24
58
2.48
49
0.18
17
1.86
14
1.67
55
0.00
1
0.00
1
0.00
1
0.00
1
0.14
73
0.05
61
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
DeepPruner_ROBtwo views0.86
18
0.50
57
1.80
30
0.01
5
0.00
1
1.57
44
0.17
51
3.48
35
0.42
5
3.31
73
0.02
3
2.09
38
0.17
16
3.01
17
0.61
27
0.00
1
0.00
1
0.00
1
0.00
1
0.12
70
0.01
13
GANet-RSSMtwo views0.88
19
0.38
38
1.45
20
0.60
103
0.00
1
1.36
19
0.02
13
4.55
76
0.69
9
0.71
15
0.32
19
1.56
22
0.61
50
4.55
44
0.64
28
0.00
1
0.00
1
0.05
109
0.00
1
0.08
54
0.02
37
ccstwo views0.89
20
0.16
13
1.12
7
0.02
15
0.02
37
1.53
41
0.24
65
6.55
114
2.18
75
0.94
24
0.22
16
1.52
21
0.36
34
2.04
15
0.51
20
0.03
101
0.02
87
0.12
118
0.05
100
0.11
69
0.06
70
PSMNet-RSSMtwo views0.90
21
0.44
47
1.35
17
0.13
52
0.00
1
1.37
22
0.03
21
4.79
89
0.85
13
1.11
32
0.49
36
2.15
41
0.35
33
4.09
30
0.73
31
0.00
1
0.00
1
0.01
88
0.00
1
0.03
22
0.01
13
DMCA-RVCcopylefttwo views0.90
21
0.64
74
3.30
67
0.03
24
0.04
49
1.30
9
0.06
36
2.65
8
0.73
10
3.21
71
0.09
8
0.99
12
0.66
56
3.14
18
0.99
41
0.01
83
0.04
97
0.02
101
0.01
85
0.07
51
0.01
13
BEATNet_4xtwo views0.97
23
0.77
89
3.36
69
0.46
95
0.01
26
1.32
13
0.09
40
4.57
78
1.26
28
1.88
46
0.22
16
1.09
18
0.42
39
1.79
13
2.09
69
0.00
1
0.00
1
0.00
1
0.00
1
0.08
54
0.06
70
cf-rtwo views0.98
24
0.51
58
2.04
40
0.25
75
0.01
26
1.40
26
0.03
21
4.47
73
1.65
61
0.95
25
0.58
38
2.63
51
0.24
24
4.19
31
0.58
25
0.00
1
0.00
1
0.00
1
0.00
1
0.06
47
0.04
56
iResNettwo views1.00
25
0.45
48
3.08
62
0.86
117
0.01
26
1.97
61
0.01
6
3.67
41
1.49
53
1.77
42
1.09
49
0.35
6
0.13
14
3.83
26
1.27
48
0.00
1
0.00
1
0.00
1
0.00
1
0.07
51
0.01
13
MLCVtwo views1.01
26
0.53
60
2.17
44
0.03
24
0.00
1
1.60
47
0.03
21
2.90
12
1.30
32
2.60
61
1.85
69
1.00
13
0.40
38
4.88
52
0.74
32
0.00
1
0.00
1
0.00
1
0.00
1
0.07
51
0.01
13
GwcNet-RSSMtwo views1.02
27
0.69
81
2.70
56
0.22
66
0.02
37
1.42
31
0.05
32
4.02
54
1.01
16
0.90
22
0.42
30
2.89
58
0.26
27
4.43
39
1.17
45
0.00
1
0.00
1
0.02
101
0.00
1
0.13
71
0.05
61
HGLStereotwo views1.03
28
0.18
16
1.50
22
0.01
5
0.01
26
1.41
28
0.01
6
3.41
30
0.82
12
1.05
29
3.09
77
2.11
39
1.24
65
4.51
41
1.18
46
0.00
1
0.00
1
0.00
1
0.00
1
0.04
35
0.03
51
ccs_robtwo views1.03
28
0.66
78
2.06
41
0.20
64
0.02
37
1.30
9
0.03
21
4.09
57
2.05
71
1.08
30
1.20
55
1.07
17
0.20
20
5.59
69
0.72
30
0.00
1
0.00
1
0.00
1
0.00
1
0.16
76
0.05
61
CC-Net-ROBtwo views1.04
30
0.57
66
2.62
54
0.50
98
0.84
114
1.36
19
0.01
6
3.40
28
1.29
31
2.28
56
0.37
27
2.21
43
0.60
49
4.43
39
0.38
13
0.00
1
0.00
1
0.00
1
0.01
85
0.01
7
0.00
1
NLCA_NET_v2_RVCtwo views1.04
30
0.55
61
2.52
51
0.56
100
1.00
117
1.37
22
0.00
1
3.43
31
1.35
35
2.24
53
0.35
25
2.18
42
0.58
48
4.28
34
0.34
9
0.00
1
0.00
1
0.00
1
0.01
85
0.00
1
0.00
1
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
iResNet_ROBtwo views1.05
32
0.76
87
2.06
41
0.07
37
0.00
1
1.35
17
0.03
21
6.80
119
2.57
87
0.79
18
1.20
55
0.94
11
0.12
12
3.81
24
0.41
14
0.00
1
0.00
1
0.00
1
0.00
1
0.08
54
0.01
13
iResNetv2_ROBtwo views1.08
33
1.12
113
4.68
93
1.09
125
0.00
1
1.38
25
0.05
32
3.83
46
1.11
22
0.80
19
1.15
52
1.02
14
0.12
12
4.52
42
0.45
17
0.00
1
0.00
1
0.00
1
0.00
1
0.33
109
0.03
51
CFNettwo views1.08
33
0.62
72
2.26
45
0.26
79
0.03
43
1.40
26
0.04
28
4.45
72
2.09
72
0.48
8
0.69
43
1.31
19
0.24
24
6.63
88
0.83
36
0.00
1
0.00
1
0.00
1
0.00
1
0.17
81
0.04
56
TDLMtwo views1.08
33
0.49
55
1.43
19
0.27
80
0.00
1
1.56
43
2.12
103
3.10
18
1.75
65
1.36
38
0.93
47
1.06
16
0.49
44
6.21
78
0.70
29
0.00
1
0.00
1
0.00
1
0.00
1
0.16
76
0.02
37
FADNet-RVC-Resampletwo views1.08
33
1.00
106
5.32
97
0.10
49
0.14
70
1.53
41
0.25
66
3.62
38
1.18
25
0.51
9
0.19
15
1.04
15
0.28
29
5.20
58
0.44
15
0.10
112
0.16
118
0.09
113
0.06
104
0.24
94
0.21
105
hitnet-ftcopylefttwo views1.09
37
0.39
40
1.45
20
0.04
30
0.00
1
1.34
16
0.22
60
3.34
26
2.25
79
1.48
39
1.11
50
4.17
75
0.37
36
4.73
47
0.79
34
0.01
83
0.00
1
0.03
105
0.00
1
0.09
64
0.01
13
PSMNet_ROBtwo views1.09
37
0.69
81
2.92
58
0.03
24
0.07
55
1.68
52
0.16
49
3.44
32
1.30
32
0.96
26
0.32
19
2.81
54
0.63
52
4.23
32
2.48
74
0.00
1
0.00
1
0.00
1
0.00
1
0.08
54
0.04
56
NVstereo2Dtwo views1.09
37
0.17
14
4.40
90
0.00
1
0.00
1
1.59
46
0.18
54
4.29
66
1.40
41
0.85
20
0.36
26
2.30
46
0.62
51
5.26
61
0.35
10
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.01
13
GANetREF_RVCpermissivetwo views1.10
40
0.95
102
2.98
59
0.06
34
0.00
1
2.23
64
0.18
54
3.95
52
2.53
85
0.56
10
0.32
19
1.89
32
0.66
56
5.03
54
0.51
20
0.00
1
0.00
1
0.00
1
0.00
1
0.05
42
0.04
56
Zhang, Feihu and Prisacariu, Victor and Yang, Ruigang and Torr, Philip HS: GA-Net: Guided Aggregation Net for End- to-end Stereo Matching. CVPR 2019
CVANet_RVCtwo views1.10
40
0.39
40
1.69
27
0.28
81
0.00
1
1.45
34
0.48
74
3.26
21
1.64
60
2.93
64
1.22
57
1.51
20
0.37
36
6.09
76
0.50
19
0.00
1
0.00
1
0.00
1
0.00
1
0.18
85
0.01
13
FADNet-RVCtwo views1.10
40
1.07
111
4.79
94
0.03
24
0.00
1
1.60
47
0.08
38
3.77
44
1.70
62
0.35
5
0.07
6
0.61
9
0.66
56
5.32
63
1.11
43
0.01
83
0.08
109
0.00
1
0.00
1
0.48
119
0.18
101
ac_64two views1.12
43
0.18
16
1.64
25
0.18
58
0.25
84
1.48
38
0.02
13
4.97
93
1.47
50
0.69
14
0.45
35
4.94
89
0.46
42
4.87
51
0.74
32
0.00
1
0.00
1
0.00
1
0.00
1
0.06
47
0.00
1
FADNet_RVCtwo views1.15
44
1.06
110
4.34
89
0.02
15
0.06
52
2.77
78
0.04
28
3.50
36
0.80
11
0.35
5
0.32
19
0.65
10
0.09
8
5.20
58
1.56
52
0.24
125
0.34
127
0.11
116
0.33
127
0.65
128
0.60
121
ETE_ROBtwo views1.16
45
0.52
59
1.71
28
0.01
5
0.02
37
1.79
58
0.22
60
4.10
58
1.32
34
4.68
92
1.18
54
2.81
54
0.08
7
3.56
20
1.12
44
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.06
70
DLCB_ROBtwo views1.17
46
0.22
24
1.28
10
0.08
41
0.00
1
1.51
40
0.23
63
3.39
27
1.52
54
3.43
77
2.00
72
3.40
69
0.31
32
4.29
35
1.82
61
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
XPNet_ROBtwo views1.17
46
0.41
44
1.68
26
0.04
30
0.01
26
1.65
51
0.17
51
3.46
34
1.39
38
3.89
81
1.01
48
1.60
24
0.53
46
4.35
38
3.07
81
0.00
1
0.00
1
0.00
1
0.00
1
0.03
22
0.02
37
FADNettwo views1.20
48
0.93
101
4.51
91
0.03
24
0.06
52
2.56
73
0.16
49
4.02
54
1.42
43
0.32
4
0.06
5
1.66
27
1.07
61
5.56
68
0.87
37
0.19
119
0.02
87
0.01
88
0.01
85
0.50
121
0.03
51
NOSS_ROBtwo views1.29
49
0.22
24
1.35
17
0.64
106
0.16
75
2.56
73
0.00
1
4.14
60
2.37
80
2.29
57
1.76
68
2.84
56
0.02
4
6.80
93
0.45
17
0.00
1
0.00
1
0.00
1
0.00
1
0.04
35
0.11
84
NCCL2two views1.30
50
0.63
73
2.31
46
0.02
15
0.03
43
1.57
44
3.41
112
3.91
51
1.39
38
3.14
70
0.34
23
3.90
73
0.30
30
3.75
22
1.19
47
0.00
1
0.00
1
0.02
101
0.00
1
0.03
22
0.05
61
DSFCAtwo views1.30
50
0.14
11
1.82
31
0.01
5
0.03
43
3.74
89
0.86
83
2.71
9
1.23
27
3.09
68
0.41
29
2.93
59
1.49
69
6.12
77
1.35
49
0.00
1
0.02
87
0.00
1
0.00
1
0.02
18
0.01
13
PA-Nettwo views1.31
52
0.59
69
3.65
77
0.06
34
0.15
73
1.44
33
0.54
78
4.08
56
2.71
89
0.78
17
1.16
53
2.32
48
0.64
53
5.96
73
1.92
64
0.01
83
0.00
1
0.00
1
0.00
1
0.04
35
0.06
70
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
LALA_ROBtwo views1.33
53
0.59
69
1.74
29
0.02
15
0.06
52
3.25
83
0.50
77
3.45
33
1.39
38
3.70
79
0.59
39
4.59
83
0.25
26
3.58
21
2.79
78
0.00
1
0.00
1
0.00
1
0.00
1
0.04
35
0.02
37
HSMtwo views1.36
54
0.40
42
1.07
6
0.01
5
0.30
94
1.73
56
0.03
21
5.89
107
1.22
26
1.83
43
1.36
61
4.78
86
1.30
67
6.61
87
0.57
24
0.00
1
0.00
1
0.00
1
0.00
1
0.02
18
0.00
1
HSM-Net_RVCpermissivetwo views1.40
55
0.07
1
1.58
23
0.00
1
0.01
26
1.88
60
0.05
32
7.24
120
1.71
64
3.63
78
3.63
85
3.21
66
0.36
34
4.25
33
0.44
15
0.01
83
0.01
74
0.00
1
0.00
1
0.01
7
0.00
1
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
StereoDRNettwo views1.48
56
0.75
84
4.58
92
0.05
33
0.10
65
2.49
71
1.10
90
4.62
79
1.47
50
3.38
76
0.62
40
2.29
45
0.65
55
4.52
42
2.89
79
0.00
1
0.03
92
0.00
1
0.01
85
0.05
42
0.07
76
RYNettwo views1.52
57
0.35
35
3.38
70
0.25
75
0.04
49
2.31
68
0.03
21
4.62
79
1.47
50
1.14
33
0.34
23
4.52
81
1.38
68
6.36
82
4.23
92
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.02
37
RASNettwo views1.53
58
0.31
32
1.97
36
0.09
46
3.20
132
1.69
54
0.06
36
3.27
22
2.12
73
0.77
16
1.29
59
5.32
96
1.69
76
6.60
86
2.17
70
0.00
1
0.00
1
0.00
1
0.00
1
0.02
18
0.02
37
CBMVpermissivetwo views1.56
59
0.24
30
1.92
34
0.31
84
0.03
43
2.90
80
4.33
119
3.79
45
1.81
66
4.98
95
1.73
66
2.70
52
1.68
75
3.81
24
0.92
39
0.00
1
0.00
1
0.00
1
0.00
1
0.05
42
0.08
78
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
PASMtwo views1.58
60
1.05
109
7.51
110
0.56
100
0.09
59
1.37
22
0.28
67
2.85
11
1.44
46
3.91
82
0.44
34
3.02
63
0.64
53
6.38
83
1.59
53
0.00
1
0.01
74
0.00
1
0.03
96
0.26
98
0.15
94
DRN-Testtwo views1.62
61
0.23
28
3.72
78
0.18
58
0.13
68
3.34
85
0.19
58
5.93
108
2.01
70
4.14
85
0.76
44
2.12
40
1.78
78
5.22
60
2.62
76
0.00
1
0.00
1
0.01
88
0.01
85
0.03
22
0.01
13
AANet_RVCtwo views1.66
62
0.88
98
3.52
74
0.06
34
0.00
1
1.22
4
0.00
1
3.25
20
2.51
84
2.21
52
4.49
89
3.65
70
0.20
20
7.10
100
3.42
85
0.50
131
0.04
97
0.00
1
0.00
1
0.08
54
0.05
61
GwcNetcopylefttwo views1.67
63
0.81
93
5.52
100
0.23
71
1.21
121
4.21
93
0.62
80
3.71
42
1.57
58
1.01
27
1.65
63
3.99
74
0.52
45
5.46
65
2.30
72
0.00
1
0.00
1
0.00
1
0.01
85
0.35
112
0.13
92
NCC-stereotwo views1.67
63
0.56
64
3.10
63
0.08
41
0.22
77
2.24
65
0.13
46
5.05
96
1.42
43
2.27
54
2.72
73
4.29
76
3.53
103
5.76
71
2.01
66
0.04
103
0.00
1
0.00
1
0.00
1
0.03
22
0.02
37
Abc-Nettwo views1.67
63
0.56
64
3.10
63
0.08
41
0.22
77
2.24
65
0.13
46
5.05
96
1.42
43
2.27
54
2.72
73
4.29
76
3.53
103
5.76
71
2.01
66
0.04
103
0.00
1
0.00
1
0.00
1
0.03
22
0.02
37
Xing Li, Yangyu Fan, Guoyun Lv, and Haoyue Ma: Area-based Correlation and Non-local Attention Network for Stereo Matching. The Visual Computer
CBMV_ROBtwo views1.70
66
0.11
8
1.98
37
0.46
95
0.00
1
2.85
79
0.91
85
4.27
64
2.21
77
4.02
84
5.00
94
2.97
61
2.32
87
5.40
64
1.35
49
0.00
1
0.00
1
0.00
1
0.00
1
0.03
22
0.06
70
DANettwo views1.74
67
0.22
24
3.06
61
0.75
111
0.44
101
2.46
70
0.35
71
3.40
28
1.41
42
4.22
86
1.56
62
5.03
91
0.56
47
6.29
80
5.00
101
0.01
83
0.00
1
0.00
1
0.01
85
0.04
35
0.08
78
RPtwo views1.76
68
0.55
61
2.59
53
0.22
66
0.60
112
1.70
55
0.98
88
4.31
68
1.99
69
3.12
69
5.04
95
4.41
79
2.96
93
4.72
46
1.90
62
0.01
83
0.00
1
0.00
1
0.01
85
0.14
73
0.01
13
Anonymous Stereotwo views1.78
69
1.37
119
7.79
111
0.42
92
0.07
55
1.49
39
3.20
110
3.31
23
1.84
67
2.75
62
1.14
51
1.84
30
0.09
8
6.89
96
3.32
84
0.00
1
0.00
1
0.00
1
0.00
1
0.09
64
0.01
13
SGM-Foresttwo views1.84
70
0.08
4
1.31
13
0.15
54
0.22
77
3.52
86
2.46
106
3.85
48
2.16
74
5.29
96
3.90
86
3.71
71
1.56
72
6.28
79
1.90
62
0.01
83
0.02
87
0.00
1
0.00
1
0.03
22
0.30
111
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
FAT-Stereotwo views1.86
71
0.19
20
3.31
68
0.17
56
0.03
43
1.78
57
0.97
86
4.76
85
2.79
90
4.22
86
4.87
92
5.31
95
2.83
92
5.06
55
0.90
38
0.00
1
0.00
1
0.00
1
0.00
1
0.04
35
0.01
13
RGCtwo views1.86
71
0.67
79
3.45
73
0.07
37
0.26
86
3.29
84
0.11
45
4.78
88
1.46
49
3.26
72
4.51
90
5.14
92
2.68
90
5.46
65
1.94
65
0.00
1
0.00
1
0.00
1
0.00
1
0.08
54
0.09
81
GANettwo views1.90
73
0.43
45
1.83
32
0.39
90
0.00
1
2.24
65
3.37
111
3.90
50
1.44
46
2.17
51
9.66
107
2.57
50
1.09
62
6.88
95
1.81
60
0.00
1
0.00
1
0.02
101
0.00
1
0.10
66
0.03
51
DISCOtwo views1.93
74
0.18
16
2.44
48
0.67
108
0.53
108
5.25
101
0.00
1
5.43
101
1.56
57
2.09
49
0.91
46
6.48
103
0.89
60
8.55
114
3.60
88
0.00
1
0.00
1
0.00
1
0.00
1
0.06
47
0.01
13
edge stereotwo views1.94
75
0.40
42
3.82
82
0.15
54
0.09
59
1.81
59
0.45
73
4.77
86
2.21
77
3.36
74
3.94
87
8.39
111
1.72
77
5.75
70
1.69
57
0.00
1
0.00
1
0.01
88
0.00
1
0.17
81
0.01
13
Nwc_Nettwo views1.94
75
0.55
61
3.79
80
0.07
37
0.08
58
2.73
75
0.04
28
6.75
118
1.11
22
2.03
48
7.92
101
4.82
87
2.11
84
4.78
48
2.01
66
0.00
1
0.00
1
0.00
1
0.00
1
0.05
42
0.01
13
S-Stereotwo views2.00
77
0.30
31
5.74
104
0.24
73
0.23
80
2.51
72
1.90
98
4.74
82
5.15
111
2.42
59
5.80
96
2.85
57
0.72
59
5.08
56
2.17
70
0.00
1
0.02
87
0.00
1
0.00
1
0.08
54
0.04
56
psmorigintwo views2.01
78
0.37
37
10.25
118
0.22
66
0.03
43
1.62
50
0.19
58
3.31
23
1.05
17
4.92
94
3.14
78
5.85
97
2.21
85
5.31
62
1.71
58
0.00
1
0.00
1
0.00
1
0.00
1
0.05
42
0.02
37
MFMNet_retwo views2.02
79
0.98
103
4.27
87
2.74
132
0.88
115
1.47
36
0.15
48
4.21
61
3.18
94
6.63
100
6.93
100
1.87
31
3.63
106
2.13
16
0.99
41
0.00
1
0.00
1
0.00
1
0.00
1
0.23
93
0.05
61
NaN_ROBtwo views2.02
79
0.58
67
2.65
55
0.29
82
0.28
93
3.19
82
5.19
124
3.84
47
2.55
86
5.31
97
1.73
66
2.30
46
1.96
82
6.78
92
3.09
82
0.01
83
0.06
105
0.03
105
0.10
111
0.10
66
0.37
116
stereogantwo views2.13
81
0.31
32
3.75
79
0.01
5
0.09
59
7.44
115
0.60
79
4.27
64
3.29
95
3.91
82
4.62
91
6.41
102
1.84
80
5.17
57
0.58
25
0.00
1
0.01
74
0.01
88
0.00
1
0.24
94
0.05
61
AF-Nettwo views2.17
82
0.65
76
3.53
75
0.00
1
0.09
59
2.02
62
0.48
74
6.44
111
1.70
62
3.36
74
10.22
109
4.95
90
2.81
91
4.34
37
2.66
77
0.00
1
0.00
1
0.00
1
0.00
1
0.13
71
0.01
13
PDISCO_ROBtwo views2.21
83
0.78
90
3.43
72
0.96
120
2.04
131
7.04
112
0.09
40
7.36
122
5.72
113
1.21
35
1.67
64
3.16
65
1.28
66
7.52
106
1.37
51
0.00
1
0.00
1
0.00
1
0.00
1
0.34
110
0.13
92
ADCReftwo views2.29
84
0.65
76
5.67
103
0.10
49
0.09
59
4.81
97
0.76
82
4.75
84
1.37
36
2.88
63
1.67
64
1.62
25
2.30
86
3.77
23
14.93
122
0.02
96
0.00
1
0.01
88
0.10
111
0.20
90
0.11
84
FBW_ROBtwo views2.36
85
0.36
36
3.16
65
0.37
89
0.14
70
4.42
96
0.18
54
6.64
115
3.05
93
2.98
66
2.85
75
4.86
88
1.11
63
11.00
126
4.53
93
0.16
118
0.05
102
0.37
128
0.09
108
0.19
89
0.78
126
DeepPrunerFtwo views2.50
86
0.87
97
14.90
124
0.17
56
0.31
95
1.41
28
0.29
68
5.62
103
9.97
119
1.02
28
0.43
33
2.01
35
1.65
74
6.85
94
3.99
89
0.00
1
0.01
74
0.07
110
0.10
111
0.24
94
0.12
91
XQCtwo views2.57
87
1.20
115
6.33
108
0.18
58
0.01
26
3.84
90
0.33
70
5.24
99
3.83
102
4.43
89
1.94
70
4.58
82
3.19
97
7.19
101
8.67
112
0.00
1
0.03
92
0.00
1
0.03
96
0.27
101
0.19
102
RTSCtwo views2.66
88
1.24
116
6.04
105
0.18
58
0.02
37
4.40
95
0.08
38
4.29
66
4.98
110
6.00
98
3.15
79
2.97
61
1.56
72
6.71
90
10.89
117
0.00
1
0.07
107
0.00
1
0.05
100
0.40
116
0.24
109
CSANtwo views2.70
89
0.75
84
3.41
71
0.18
58
0.12
66
4.18
92
4.05
116
4.11
59
4.36
106
6.03
99
6.24
99
4.60
84
4.74
110
7.29
103
3.43
86
0.02
96
0.03
92
0.01
88
0.02
95
0.18
85
0.22
106
ADCP+two views2.87
90
0.98
103
8.75
115
0.07
37
0.16
75
5.93
107
2.90
109
4.00
53
1.84
67
1.52
40
0.27
18
3.81
72
3.22
99
6.05
75
17.38
125
0.00
1
0.00
1
0.00
1
0.00
1
0.14
73
0.30
111
aanetorigintwo views2.87
90
0.73
83
8.74
114
0.23
71
0.23
80
1.61
49
2.07
101
1.76
3
1.10
21
14.21
123
10.25
110
2.93
59
3.38
101
4.98
53
4.06
90
0.10
112
0.05
102
0.09
113
0.11
114
0.57
124
0.15
94
PWC_ROBbinarytwo views2.90
92
1.01
107
6.12
106
0.53
99
0.57
111
1.30
9
0.40
72
3.85
48
4.42
108
7.37
104
15.40
120
3.03
64
1.55
71
6.31
81
5.74
107
0.00
1
0.00
1
0.00
1
0.00
1
0.46
117
0.05
61
SHDtwo views3.01
93
1.02
108
5.52
100
0.45
94
0.39
98
4.12
91
0.17
51
6.68
116
12.04
122
7.20
103
3.36
82
4.40
78
2.58
89
4.69
45
7.17
109
0.01
83
0.04
97
0.00
1
0.03
96
0.22
92
0.17
99
PWCDC_ROBbinarytwo views3.06
94
1.38
120
3.81
81
0.14
53
0.00
1
3.14
81
0.02
13
4.40
70
12.35
123
1.08
30
23.69
129
2.01
35
2.01
83
3.87
27
2.43
73
0.24
125
0.00
1
0.00
1
0.00
1
0.61
126
0.08
78
SPS-STEREOcopylefttwo views3.08
95
0.49
55
2.49
50
0.21
65
0.24
83
3.67
88
1.24
91
5.80
106
2.43
83
10.68
115
8.00
102
9.48
116
3.59
105
8.11
111
5.08
102
0.01
83
0.01
74
0.00
1
0.00
1
0.04
35
0.11
84
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
NVStereoNet_ROBtwo views3.25
96
0.46
49
2.71
57
0.34
87
0.56
110
2.36
69
0.87
84
4.56
77
4.47
109
4.55
90
14.09
118
11.93
118
3.23
100
9.79
122
3.14
83
0.43
129
0.03
92
0.53
131
0.31
126
0.36
114
0.38
117
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
ADCLtwo views3.30
97
0.81
93
6.76
109
0.25
75
0.25
84
6.93
110
4.65
120
4.79
89
2.62
88
4.61
91
3.36
82
5.22
93
4.55
109
5.54
67
14.91
121
0.05
106
0.01
74
0.11
116
0.13
115
0.28
103
0.10
82
MFN_U_SF_DS_RVCtwo views3.30
97
2.61
128
7.84
112
0.09
46
0.26
86
12.92
125
7.62
130
4.25
62
3.64
100
2.95
65
3.99
88
4.61
85
1.22
64
8.36
112
1.77
59
0.59
132
1.15
134
0.04
107
0.85
132
0.48
119
0.67
124
SAMSARAtwo views3.32
99
1.07
111
6.25
107
0.86
117
0.31
95
7.55
116
3.87
115
5.26
100
4.20
105
7.43
105
3.35
81
10.04
117
3.83
108
6.98
99
4.83
98
0.00
1
0.14
115
0.00
1
0.05
100
0.17
81
0.22
106
DPSNettwo views3.38
100
0.60
71
9.55
116
0.61
104
0.46
103
5.10
100
1.44
95
11.17
130
3.92
103
2.57
60
0.88
45
5.90
98
9.68
124
7.63
107
7.25
110
0.08
108
0.16
118
0.01
88
0.07
105
0.34
110
0.17
99
STTStereo_v2two views3.49
101
0.47
52
4.03
85
1.61
126
1.91
126
10.78
123
1.92
99
3.06
16
1.26
28
8.59
109
12.06
112
6.13
100
7.82
118
4.78
48
4.77
96
0.08
108
0.12
113
0.07
110
0.15
116
0.16
76
0.11
84
G-Nettwo views3.49
101
0.47
52
4.03
85
1.61
126
1.91
126
10.78
123
1.92
99
3.06
16
1.26
28
8.59
109
12.06
112
6.13
100
7.82
118
4.78
48
4.77
96
0.08
108
0.12
113
0.07
110
0.15
116
0.16
76
0.11
84
MDST_ROBtwo views3.50
103
0.07
1
3.94
83
2.66
131
1.46
124
12.94
126
0.97
86
7.32
121
2.42
82
14.82
125
8.08
103
2.74
53
1.51
70
8.44
113
2.48
74
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.11
84
pmcnntwo views3.61
104
0.68
80
5.38
99
0.61
104
1.46
124
2.75
77
1.26
92
5.04
95
1.57
58
8.77
111
11.67
111
18.69
128
2.38
88
6.03
74
5.61
106
0.00
1
0.07
107
0.00
1
0.00
1
0.21
91
0.07
76
ADCPNettwo views3.63
105
0.82
95
13.20
122
0.08
41
0.43
100
7.02
111
2.23
104
4.77
86
2.41
81
4.31
88
1.96
71
7.57
109
5.16
111
6.45
85
13.80
120
0.02
96
0.75
132
0.04
107
0.90
133
0.16
76
0.53
120
SuperBtwo views3.70
106
1.77
122
15.73
125
0.31
84
0.05
51
4.37
94
1.67
97
3.58
37
11.57
120
2.36
58
1.32
60
2.03
37
1.91
81
6.95
98
9.01
113
0.05
106
0.03
92
0.10
115
0.08
106
11.00
134
0.10
82
MFN_U_SF_RVCtwo views3.76
107
1.77
122
9.85
117
0.84
112
0.27
89
9.50
120
0.49
76
8.07
123
3.41
96
4.81
93
3.28
80
9.40
115
3.49
102
7.76
109
8.31
111
0.41
128
0.54
129
0.52
130
0.80
131
0.89
129
0.84
127
AnyNet_C32two views3.93
108
2.58
127
12.07
121
0.39
90
0.49
106
5.50
103
6.45
128
4.44
71
2.19
76
6.86
102
2.94
76
5.28
94
3.20
98
6.63
88
18.37
127
0.08
108
0.04
97
0.20
122
0.09
108
0.57
124
0.19
102
ADCMidtwo views3.95
109
1.34
117
10.91
120
0.18
58
0.38
97
5.02
99
1.27
93
4.87
91
2.94
92
11.15
116
3.60
84
6.76
106
5.21
113
6.40
84
16.96
124
0.12
115
0.08
109
0.55
132
0.46
128
0.53
123
0.19
102
SGM_RVCbinarytwo views4.07
110
0.38
38
1.98
37
0.84
112
0.26
86
6.20
108
2.37
105
6.45
112
3.57
98
11.56
118
9.27
106
15.55
125
6.78
115
10.07
123
4.60
95
0.22
120
0.25
123
0.20
122
0.23
123
0.26
98
0.34
114
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
SGM+DAISYtwo views4.39
111
1.34
117
5.18
96
1.01
122
0.88
115
5.61
105
2.86
108
4.88
92
3.53
97
13.29
120
13.95
117
12.66
120
7.55
116
7.98
110
5.39
105
0.23
124
0.18
121
0.15
120
0.24
124
0.29
104
0.64
122
WCMA_ROBtwo views4.54
112
0.48
54
3.19
66
0.49
97
0.63
113
5.38
102
2.57
107
4.70
81
3.77
101
14.17
122
19.02
123
15.47
124
9.03
121
6.73
91
4.85
99
0.01
83
0.05
102
0.01
88
0.01
85
0.08
54
0.16
96
SANettwo views4.63
113
0.89
99
5.34
98
0.30
83
0.09
59
7.43
114
3.75
114
6.53
113
16.16
127
6.79
101
13.29
115
12.44
119
7.74
117
7.38
104
4.11
91
0.00
1
0.01
74
0.00
1
0.00
1
0.10
66
0.31
113
MSMD_ROBtwo views4.94
114
0.43
45
2.58
52
0.12
51
0.01
26
8.43
119
1.09
89
4.25
62
4.11
104
13.11
119
26.36
131
13.95
123
13.88
130
7.45
105
2.92
80
0.10
112
0.01
74
0.00
1
0.00
1
0.01
7
0.02
37
FCDSN-DCtwo views5.04
115
0.15
12
1.83
32
0.22
66
0.44
101
5.60
104
1.66
96
5.22
98
4.37
107
14.49
124
21.86
127
20.65
129
11.88
128
7.66
108
4.55
94
0.02
96
0.00
1
0.00
1
0.00
1
0.03
22
0.05
61
Dominik Hirner, Friedrich Fraundorfer: FCDSN-DC: An accurate but lightweight end-to-end trainable neural network for stereo estimation with depth completion.
ADCStwo views5.09
116
2.15
125
16.58
126
0.34
87
0.07
55
5.91
106
4.26
118
6.73
117
6.99
116
9.44
113
6.09
98
6.57
104
5.17
112
9.54
120
21.38
132
0.00
1
0.00
1
0.00
1
0.00
1
0.36
114
0.23
108
MSC_U_SF_DS_RVCtwo views5.12
117
2.33
126
8.42
113
0.33
86
0.27
89
24.60
135
8.84
132
5.50
102
9.75
118
3.82
80
8.16
104
8.53
112
1.82
79
9.44
119
5.18
103
1.02
133
1.65
135
0.33
127
0.47
129
1.06
131
0.90
129
MeshStereopermissivetwo views5.78
118
0.75
84
3.04
60
0.25
75
0.14
70
7.88
118
2.09
102
9.64
126
5.18
112
20.48
129
19.06
124
22.80
130
9.54
123
9.64
121
4.88
100
0.00
1
0.00
1
0.00
1
0.00
1
0.29
104
0.03
51
C. Zhang, Z. Li, Y. Cheng, R. Cai, H. Chao, Y. Rui: MeshStereo: A Global Stereo Model with Mesh Alignment Regularization for View Interpolation. ICCV 2015
PVDtwo views5.81
119
0.99
105
5.56
102
0.72
110
0.51
107
6.72
109
0.18
54
9.90
128
23.69
132
11.45
117
19.26
125
8.18
110
9.46
122
9.24
117
9.49
114
0.02
96
0.16
118
0.00
1
0.08
106
0.18
85
0.40
118
FC-DCNNcopylefttwo views6.09
120
0.18
16
2.35
47
0.42
92
0.53
108
7.80
117
1.42
94
8.14
124
7.07
117
19.05
127
22.10
128
22.85
131
13.84
129
9.42
118
6.44
108
0.00
1
0.01
74
0.00
1
0.00
1
0.01
7
0.06
70
EDNetEfficienttwo views6.54
121
1.17
114
33.15
130
0.66
107
1.96
128
2.07
63
3.72
113
2.80
10
14.85
126
20.02
128
9.69
108
6.59
105
10.19
125
7.19
101
15.49
123
0.03
101
0.10
112
0.01
88
0.20
119
0.62
127
0.29
110
AnyNet_C01two views6.59
122
4.31
131
36.24
131
0.96
120
0.46
103
9.96
121
5.74
125
6.02
110
3.63
99
7.59
106
8.56
105
8.83
113
3.14
96
13.74
128
20.83
131
0.12
115
0.08
109
0.25
125
0.09
108
1.01
130
0.34
114
LSMtwo views7.38
123
1.68
121
22.24
128
3.99
133
40.72
136
3.54
87
4.79
123
4.74
82
6.65
114
10.25
114
13.51
116
4.49
80
3.81
107
6.91
97
5.31
104
0.00
1
0.04
97
0.00
1
0.00
1
0.26
98
14.60
135
ELAS_RVCcopylefttwo views7.69
124
0.80
91
4.90
95
1.06
123
1.02
118
10.00
122
9.43
133
9.75
127
14.69
125
21.83
130
20.27
126
17.88
126
16.41
132
13.01
127
10.44
116
0.22
120
0.55
130
0.28
126
0.19
118
0.35
112
0.65
123
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
DispFullNettwo views7.91
125
23.96
136
13.67
123
14.71
135
7.88
134
4.83
98
0.04
28
5.01
94
2.89
91
8.86
112
4.96
93
6.11
99
10.72
126
10.18
124
3.51
87
2.44
134
0.33
126
15.15
135
4.65
135
12.33
135
5.95
134
ELAScopylefttwo views8.05
126
0.76
87
4.02
84
0.95
119
1.02
118
14.85
129
6.99
129
12.80
132
11.73
121
22.43
131
27.75
132
17.88
126
15.19
131
10.96
125
11.31
118
0.22
120
0.56
131
0.23
124
0.22
121
0.50
121
0.71
125
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
PWCKtwo views8.72
127
5.37
132
16.61
127
5.74
134
0.12
66
19.88
131
11.41
134
11.31
131
13.42
124
14.07
121
12.84
114
13.20
122
8.05
120
15.44
129
10.25
115
5.95
135
0.22
122
2.09
134
0.05
100
7.50
133
0.92
130
RTSAtwo views9.75
128
2.69
129
65.25
134
0.85
114
1.99
129
14.36
127
4.75
121
5.69
104
17.29
129
7.97
107
19.00
121
6.83
107
3.02
94
26.15
133
18.62
129
0.00
1
0.15
116
0.00
1
0.00
1
0.29
104
0.16
96
RTStwo views9.75
128
2.69
129
65.25
134
0.85
114
1.99
129
14.36
127
4.75
121
5.69
104
17.29
129
7.97
107
19.00
121
6.83
107
3.02
94
26.15
133
18.62
129
0.00
1
0.15
116
0.00
1
0.00
1
0.29
104
0.16
96
EDNetEfficientorigintwo views10.31
130
1.90
124
98.96
139
0.22
66
0.23
80
2.73
75
0.71
81
4.50
74
16.94
128
18.72
126
15.36
119
12.74
121
10.73
127
8.96
115
11.94
119
0.01
83
0.01
74
0.00
1
0.00
1
0.25
97
1.31
133
MADNet+two views10.71
131
14.57
134
70.07
136
0.58
102
0.47
105
19.89
132
4.20
117
11.05
129
6.87
115
3.03
67
5.99
97
9.01
114
6.35
114
31.95
136
28.31
135
0.22
120
0.26
125
0.01
88
0.04
99
0.46
117
0.87
128
SGM-ForestMtwo views12.56
132
0.80
91
4.31
88
1.07
124
0.41
99
21.56
133
8.57
131
17.42
133
19.67
131
29.93
132
28.49
134
45.83
135
27.22
134
26.05
132
18.40
128
0.14
117
0.25
123
0.18
121
0.28
125
0.18
85
0.42
119
LE_ROBtwo views12.79
133
0.64
74
10.26
119
2.13
129
1.29
122
7.25
113
5.81
126
9.09
125
49.27
136
57.55
136
24.58
130
23.68
132
31.13
136
9.00
116
23.57
134
0.04
103
0.06
105
0.13
119
0.20
119
0.08
54
0.11
84
MANEtwo views14.02
134
0.85
96
3.53
75
2.13
129
3.75
133
18.44
130
6.14
127
22.67
135
27.90
134
33.01
134
41.08
136
41.26
134
30.90
135
23.39
130
17.99
126
0.47
130
0.76
133
0.91
133
4.04
134
0.17
81
1.00
131
BEATNet-Init1two views17.91
135
7.45
133
36.82
132
1.88
128
1.38
123
37.29
136
14.71
135
21.90
134
24.20
133
38.07
135
36.59
135
50.41
136
32.00
137
29.42
135
21.83
133
0.33
127
0.39
128
0.47
129
0.54
130
1.50
132
1.07
132
DPSimNet_ROBtwo views25.03
136
19.56
135
24.93
129
24.36
136
17.85
135
22.25
134
25.93
136
28.52
136
29.10
135
32.61
133
28.25
133
32.87
133
26.42
133
24.51
131
40.42
136
16.75
136
13.93
136
25.19
136
22.38
136
19.96
136
24.75
136
MADNet++two views47.78
137
34.15
137
37.11
133
42.94
137
41.23
137
64.94
137
30.63
137
60.79
137
53.26
137
72.08
137
61.65
137
57.68
137
54.96
138
58.03
137
59.24
137
43.01
137
36.17
137
36.07
137
22.50
137
47.50
137
41.67
137
MEDIAN_ROBtwo views96.83
138
99.41
139
98.66
138
94.75
138
94.23
138
93.08
138
90.54
138
95.61
138
94.75
138
97.65
140
97.73
138
98.30
138
96.57
139
94.51
138
93.74
138
99.78
143
98.24
138
99.99
140
99.89
138
99.48
140
99.67
140
DPSM_ROBtwo views99.17
139
100.00
143
100.00
142
96.28
139
98.84
139
100.00
140
100.00
140
100.00
141
100.00
139
100.00
141
100.00
139
100.00
141
100.00
140
100.00
142
100.00
142
93.66
138
100.00
139
100.00
141
100.00
139
96.31
138
98.27
138
DPSMtwo views99.17
139
100.00
143
100.00
142
96.28
139
98.84
139
100.00
140
100.00
140
100.00
141
100.00
139
100.00
141
100.00
139
100.00
141
100.00
140
100.00
142
100.00
142
93.66
138
100.00
139
100.00
141
100.00
139
96.31
138
98.27
138
AVERAGE_ROBtwo views99.23
141
99.81
140
97.56
137
100.00
144
100.00
141
96.24
139
91.02
139
100.00
141
100.00
139
100.00
141
100.00
139
100.00
141
100.00
140
99.98
141
99.97
139
100.00
144
100.00
139
100.00
141
100.00
139
100.00
144
100.00
143
DGTPSM_ROBtwo views99.45
142
99.94
141
99.98
140
96.58
141
100.00
141
100.00
140
100.00
140
99.81
139
100.00
139
95.86
138
100.00
139
99.32
139
100.00
140
99.79
139
99.99
140
98.34
140
100.00
139
99.86
138
100.00
139
99.52
141
99.99
142
DPSMNet_ROBtwo views99.46
143
99.94
141
99.98
140
96.59
142
100.00
141
100.00
140
100.00
140
99.81
139
100.00
139
95.87
139
100.00
139
99.32
139
100.00
140
99.79
139
99.99
140
98.46
141
100.00
139
99.86
138
100.00
139
99.52
141
100.00
143
LSM0two views99.94
144
100.00
143
100.00
142
99.84
143
100.00
141
100.00
140
100.00
140
100.00
141
100.00
139
100.00
141
100.00
139
100.00
141
100.00
140
100.00
142
100.00
142
99.22
142
100.00
139
100.00
141
100.00
139
99.99
143
99.83
141
MSMDNettwo views0.42
39
ASD4two views67.10
138