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
CREStereotwo views0.92
1
0.53
3
1.45
1
0.61
1
0.34
1
4.42
109
0.40
1
2.21
2
1.54
3
0.94
1
1.12
3
0.87
2
0.87
2
1.12
1
0.72
1
0.28
3
0.19
2
0.22
1
0.15
1
0.22
1
0.19
1
PMTNettwo views0.99
2
0.48
1
1.73
2
0.62
2
0.56
2
4.20
99
0.43
2
1.88
1
1.58
5
1.00
2
1.05
1
0.78
1
0.98
5
1.50
8
0.85
2
1.21
124
0.20
4
0.23
2
0.15
1
0.22
1
0.19
1
RAFT-Stereopermissivetwo views1.20
3
0.58
7
1.81
4
0.99
8
0.67
9
3.78
58
0.90
24
4.09
80
1.51
1
2.16
37
1.71
23
1.13
4
0.94
3
1.40
6
0.91
4
0.30
6
0.24
16
0.26
6
0.21
5
0.26
3
0.25
4
Lahav Lipson, Zachary Teed, and Jia Deng: RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching. 3DV
R-Stereo Traintwo views1.20
3
0.58
7
1.81
4
0.99
8
0.67
9
3.78
58
0.90
24
4.09
80
1.51
1
2.16
37
1.71
23
1.13
4
0.94
3
1.40
6
0.91
4
0.30
6
0.24
16
0.26
6
0.21
5
0.26
3
0.25
4
DPM-Stereotwo views1.24
5
0.69
15
1.77
3
0.65
3
0.61
8
4.29
104
0.58
4
4.44
114
2.53
65
1.98
25
1.33
6
0.91
3
0.86
1
1.25
3
1.25
6
0.36
28
0.23
13
0.32
16
0.23
11
0.32
10
0.29
11
AdaStereotwo views1.45
6
0.83
32
2.34
9
1.28
51
0.81
17
3.21
12
1.14
47
4.20
92
2.46
59
2.05
31
1.77
30
2.48
28
1.18
8
1.71
11
1.44
8
0.43
49
0.25
20
0.43
46
0.28
21
0.44
29
0.34
17
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.
ACVNettwo views1.46
7
0.64
13
3.56
19
0.93
5
0.85
22
4.22
100
0.80
13
2.85
5
2.08
21
1.78
10
1.88
41
3.10
59
1.46
24
1.33
4
1.89
35
0.29
4
0.26
24
0.30
12
0.26
16
0.39
18
0.27
8
acv_fttwo views1.55
8
0.64
13
3.29
16
1.14
24
2.01
118
2.93
10
0.97
36
3.81
40
2.08
21
2.22
40
1.88
41
3.10
59
1.66
42
1.33
4
2.22
51
0.29
4
0.26
24
0.30
12
0.26
16
0.39
18
0.27
8
Gwc-CoAtRStwo views1.58
9
0.57
5
11.69
45
1.01
10
0.60
7
2.53
2
0.64
7
2.36
3
1.56
4
1.80
13
1.47
9
2.74
39
1.18
8
1.12
1
0.90
3
0.27
2
0.21
7
0.25
4
0.22
9
0.27
5
0.23
3
ccstwo views1.61
10
0.55
4
2.29
8
1.27
50
0.80
14
3.62
45
0.75
12
4.22
96
2.77
75
1.94
22
1.60
13
2.54
32
1.46
24
3.60
20
1.72
22
0.40
41
0.60
101
0.38
30
0.39
52
0.45
33
0.77
90
DMCAtwo views1.72
11
0.74
23
6.63
24
1.14
24
0.96
33
3.98
77
0.90
24
3.75
37
2.76
74
2.83
72
1.58
12
2.35
23
1.18
8
1.58
9
1.98
40
0.40
41
0.27
27
0.42
43
0.26
16
0.36
15
0.38
23
StereoDRNet-Refinedtwo views1.77
12
0.81
28
2.22
7
1.21
36
0.76
13
3.26
16
0.71
9
3.82
41
2.05
19
2.29
44
2.05
54
2.98
56
1.44
19
7.15
47
2.27
53
0.30
6
0.23
13
0.42
43
0.33
30
0.45
33
0.64
70
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
MLCVtwo views1.87
13
0.91
36
8.59
29
1.10
19
0.57
5
3.59
43
0.53
3
3.70
34
2.20
30
2.80
70
2.25
71
2.00
12
1.42
17
4.13
24
1.86
33
0.30
6
0.20
4
0.32
16
0.21
5
0.33
12
0.36
20
DSFCAtwo views1.89
14
0.69
15
3.14
14
1.10
19
1.18
54
3.99
79
1.79
81
3.50
21
2.16
28
3.10
83
1.64
15
2.91
54
2.50
77
5.38
33
2.26
52
0.43
49
0.38
55
0.37
27
0.36
39
0.50
40
0.42
33
iResNet_ROBtwo views1.89
14
1.02
48
8.60
30
1.39
77
0.98
40
3.72
54
0.66
8
3.84
42
2.65
69
1.88
18
2.05
54
1.98
11
1.27
12
4.52
26
1.67
18
0.30
6
0.20
4
0.28
9
0.18
3
0.31
9
0.40
28
CFNet_RVCtwo views1.90
16
0.95
41
3.56
19
1.07
13
1.54
84
4.09
86
0.90
24
2.85
5
1.82
10
1.78
10
1.70
18
2.75
41
1.48
27
9.33
65
1.72
22
0.33
17
0.30
32
0.61
95
0.36
39
0.52
43
0.39
24
CFNet-ftpermissivetwo views1.90
16
0.95
41
3.56
19
1.07
13
1.54
84
4.09
86
0.90
24
2.85
5
1.82
10
1.78
10
1.70
18
2.75
41
1.48
27
9.33
65
1.72
22
0.33
17
0.30
32
0.61
95
0.36
39
0.52
43
0.39
24
DN-CSS_ROBtwo views1.91
18
1.55
91
11.13
43
1.82
110
0.88
24
3.99
79
0.61
6
4.01
68
1.67
6
1.64
7
1.89
43
1.39
7
1.06
6
3.23
15
1.50
10
0.31
11
0.19
2
0.44
49
0.29
23
0.37
17
0.31
14
BEATNet_4xtwo views1.92
19
1.36
79
8.75
31
1.20
35
0.56
2
3.46
30
1.03
39
4.00
67
2.43
54
2.25
43
1.70
18
2.10
18
1.59
36
2.57
14
3.05
76
0.37
29
0.22
10
0.47
57
0.29
23
0.60
63
0.40
28
LALA_ROBtwo views1.95
20
1.48
88
3.11
13
1.05
12
0.98
40
3.99
79
1.82
83
4.44
114
2.28
40
3.31
90
1.85
39
2.76
46
1.67
44
4.04
23
3.23
78
0.46
59
0.41
63
0.55
83
0.52
82
0.60
63
0.57
59
PSMNet-RSSMtwo views1.96
21
0.77
25
3.26
15
1.29
54
1.31
68
4.19
95
0.96
35
3.68
30
1.89
13
2.05
31
1.80
34
2.60
36
1.70
46
9.84
72
1.81
30
0.35
26
0.29
30
0.35
25
0.37
47
0.42
26
0.35
18
GANet-RSSMtwo views1.98
22
0.62
10
3.49
17
1.69
103
1.23
60
4.22
100
0.91
30
3.31
15
1.86
12
1.81
14
1.74
28
2.34
22
1.82
55
10.92
96
1.73
27
0.34
21
0.29
30
0.31
14
0.28
21
0.39
18
0.40
28
HGLStereotwo views2.03
23
0.74
23
7.50
25
1.23
40
1.56
91
3.62
45
0.95
34
3.78
39
1.89
13
2.03
29
2.20
67
3.12
62
2.14
65
5.68
40
2.04
43
0.37
29
0.31
36
0.32
16
0.29
23
0.40
22
0.45
36
DeepPruner_ROBtwo views2.05
24
1.09
57
10.50
40
1.01
10
1.20
57
4.39
107
0.91
30
3.14
11
1.72
8
2.60
58
1.54
10
2.78
47
1.27
12
4.52
26
1.72
22
0.51
71
0.40
60
0.37
27
0.34
31
0.48
37
0.52
49
DLCB_ROBtwo views2.05
24
0.95
41
2.37
10
1.24
42
1.02
44
2.92
9
1.30
54
3.68
30
2.30
42
2.91
74
2.17
64
3.32
65
1.52
33
10.37
84
2.50
59
0.40
41
0.33
39
0.40
38
0.36
39
0.41
25
0.47
39
HITNettwo views2.07
26
0.82
31
14.77
62
1.34
65
0.56
2
3.47
32
0.80
13
4.07
77
2.35
46
2.16
37
1.61
14
1.64
9
1.44
19
2.09
12
2.56
62
0.26
1
0.18
1
0.41
42
0.26
16
0.45
33
0.25
4
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
ETE_ROBtwo views2.07
26
1.40
84
3.01
12
1.09
15
0.92
30
3.94
71
1.53
67
4.25
98
2.23
33
3.31
90
2.06
56
2.35
23
1.52
33
8.74
57
2.07
46
0.44
53
0.36
48
0.51
72
0.48
74
0.54
50
0.62
66
DISCOtwo views2.25
28
0.73
20
10.06
37
1.76
108
1.12
49
3.47
32
1.25
53
3.24
13
2.26
37
2.65
60
1.96
45
4.53
104
1.92
60
4.17
25
3.61
80
0.32
13
0.25
20
0.31
14
0.29
23
0.66
77
0.41
32
FENettwo views2.25
28
0.50
2
20.29
113
1.35
68
0.67
9
4.01
82
0.60
5
3.95
63
1.99
15
1.96
23
1.11
2
2.43
27
1.43
18
1.60
10
1.39
7
0.32
13
0.22
10
0.33
21
0.26
16
0.29
7
0.39
24
NOSS_ROBtwo views2.27
30
0.73
20
3.00
11
1.71
105
1.00
42
3.90
67
0.97
36
4.29
102
2.81
82
2.94
77
2.19
66
3.88
83
1.19
11
11.77
110
1.62
13
0.59
83
0.54
90
0.56
84
0.52
82
0.56
56
0.54
53
XPNet_ROBtwo views2.27
30
1.11
60
4.10
22
1.36
72
0.97
37
3.94
71
1.63
73
4.07
77
2.14
25
2.82
71
2.00
48
2.32
21
1.73
48
10.81
91
3.39
79
0.51
71
0.47
81
0.48
62
0.40
59
0.51
41
0.56
56
TDLMtwo views2.29
32
1.08
55
4.61
23
1.29
54
1.01
43
4.26
103
2.17
93
3.75
37
2.79
78
2.31
46
1.98
47
2.06
17
1.59
36
12.45
119
1.72
22
0.51
71
0.28
29
0.52
78
0.35
34
0.56
56
0.50
45
CFNettwo views2.34
33
1.10
58
8.96
32
1.34
65
1.24
61
4.09
86
0.73
11
4.04
72
2.77
75
1.62
6
1.87
40
2.27
20
1.45
22
11.41
97
1.82
31
0.31
11
0.24
16
0.43
46
0.36
39
0.43
28
0.33
15
STTStereotwo views2.34
33
0.95
41
14.93
67
1.43
79
1.54
84
3.86
66
0.86
19
3.65
27
2.44
56
2.15
36
1.55
11
2.49
30
1.59
36
5.14
30
1.44
8
0.38
35
0.39
59
0.47
57
0.51
79
0.52
43
0.51
48
ac_64two views2.35
35
0.62
10
9.77
36
1.32
57
1.71
103
3.27
18
0.93
32
3.48
19
2.50
62
1.74
9
1.73
27
3.91
84
1.77
50
10.21
80
1.86
33
0.40
41
0.31
36
0.33
21
0.30
27
0.34
13
0.39
24
HSMtwo views2.35
35
0.81
28
2.21
6
1.14
24
1.55
88
3.44
26
1.03
39
3.99
66
2.25
35
2.29
44
2.11
60
8.04
123
3.05
90
11.60
101
1.71
20
0.33
17
0.22
10
0.28
9
0.23
11
0.32
10
0.35
18
hitnet-ftcopylefttwo views2.37
37
0.73
20
8.34
27
0.96
6
0.89
26
3.77
57
0.80
13
3.55
24
2.58
68
2.24
42
2.04
51
3.65
78
1.49
29
11.71
108
1.89
35
0.44
53
0.53
89
0.44
49
0.35
34
0.52
43
0.48
40
DMCA-RVCcopylefttwo views2.45
38
1.22
70
18.29
97
1.23
40
1.14
51
2.49
1
0.85
18
3.29
14
1.68
7
3.02
79
1.41
8
2.00
12
1.53
35
6.39
43
2.00
41
0.38
35
0.30
32
0.44
49
0.35
34
0.45
33
0.44
34
FADNettwo views2.49
39
1.76
101
13.17
51
1.17
28
0.95
32
3.27
18
0.88
23
3.94
62
2.27
39
1.40
4
1.29
5
2.51
31
2.06
62
9.93
76
1.94
37
0.43
49
0.45
76
0.40
38
0.43
65
1.06
110
0.53
50
FADNet-RVCtwo views2.51
40
2.27
112
14.03
55
1.09
15
0.81
17
3.54
40
1.10
43
4.13
84
2.48
60
1.39
3
1.22
4
1.85
10
1.77
50
9.78
71
2.06
45
0.38
35
0.37
53
0.40
38
0.40
59
0.67
78
0.48
40
iResNettwo views2.52
41
0.94
40
21.17
122
1.84
112
0.72
12
3.75
56
0.81
16
4.14
87
2.55
67
2.33
47
2.03
49
1.35
6
1.44
19
3.27
16
2.16
49
0.35
26
0.23
13
0.28
9
0.22
9
0.44
29
0.45
36
NCCL2two views2.53
42
1.35
78
10.90
41
1.15
27
0.97
37
3.53
38
2.56
99
3.65
27
2.23
33
2.76
66
1.80
34
2.41
26
1.66
42
10.35
83
2.07
46
0.44
53
0.40
60
0.57
85
0.53
86
0.59
60
0.65
73
NVstereo2Dtwo views2.54
43
0.90
34
11.10
42
1.32
57
1.47
81
3.58
42
1.38
58
4.13
84
2.45
57
1.90
20
1.71
23
2.86
52
1.76
49
10.86
92
1.66
17
0.60
84
0.43
69
0.53
79
0.41
61
0.78
92
0.94
106
NLCA_NET_v2_RVCtwo views2.54
43
1.04
50
16.73
85
1.53
91
2.00
117
3.96
74
0.87
22
3.87
48
2.31
43
2.54
54
1.67
17
2.85
51
1.63
40
5.55
36
1.62
13
0.46
59
0.33
39
0.38
30
0.38
49
0.53
49
0.62
66
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
ccs_robtwo views2.54
43
1.14
62
14.21
58
1.35
68
0.96
33
3.98
77
0.71
9
3.86
46
2.88
86
2.04
30
2.07
57
2.05
16
1.29
15
10.60
87
1.71
20
0.32
13
0.25
20
0.38
30
0.35
34
0.39
18
0.33
15
CC-Net-ROBtwo views2.55
46
1.05
51
16.77
87
1.50
86
1.94
116
3.97
76
0.90
24
3.85
44
2.25
35
2.55
55
1.70
18
2.82
49
1.65
41
5.67
39
1.65
15
0.47
62
0.33
39
0.38
30
0.39
52
0.54
50
0.64
70
CVANet_RVCtwo views2.55
46
1.10
58
9.13
33
1.46
81
1.15
52
3.96
74
1.53
67
3.91
58
2.50
62
3.04
80
2.04
51
2.60
36
1.72
47
12.46
120
1.50
10
0.49
65
0.30
32
0.51
72
0.41
61
0.63
68
0.50
45
RASNettwo views2.56
48
0.70
17
10.06
37
1.60
96
2.25
125
3.67
50
1.00
38
3.38
18
3.14
95
1.81
14
2.11
60
4.60
106
2.30
69
10.25
82
2.61
63
0.37
29
0.24
16
0.25
4
0.24
15
0.28
6
0.26
7
cf-rtwo views2.57
49
0.79
27
16.09
75
1.48
84
1.35
72
4.19
95
0.93
32
3.68
30
2.36
48
1.97
24
1.81
36
3.42
67
1.45
22
8.24
52
1.83
32
0.32
13
0.26
24
0.33
21
0.30
27
0.34
13
0.37
22
DANettwo views2.63
50
1.07
54
8.10
26
1.83
111
1.70
102
3.25
15
1.16
48
4.31
107
2.38
51
3.40
93
2.18
65
4.08
91
1.80
53
9.77
70
3.89
83
0.52
76
0.55
92
0.59
88
0.44
66
0.63
68
0.86
98
iResNetv2_ROBtwo views2.65
51
5.00
124
19.28
106
2.06
123
1.09
47
3.83
64
1.13
46
4.14
87
2.09
23
1.85
16
2.03
49
2.01
14
1.51
31
3.58
18
1.58
12
0.34
21
0.21
7
0.32
16
0.21
5
0.44
29
0.29
11
FADNet_RVCtwo views2.65
51
2.28
113
16.51
81
0.98
7
0.92
30
4.09
86
0.83
17
4.26
99
1.73
9
1.46
5
1.78
32
1.60
8
1.50
30
8.88
59
2.39
55
0.47
62
0.38
55
0.44
49
0.58
94
0.89
100
0.97
107
GwcNet-RSSMtwo views2.66
53
1.05
51
15.91
74
1.22
39
1.19
55
4.12
91
1.12
45
3.84
42
2.02
16
1.93
21
1.79
33
3.12
62
1.46
24
10.37
84
2.05
44
0.33
17
0.27
27
0.35
25
0.30
27
0.40
22
0.36
20
PASMtwo views2.67
54
2.05
109
9.71
35
1.73
107
1.37
73
2.53
2
1.52
66
3.10
10
2.17
29
2.79
69
1.84
37
2.90
53
1.85
58
11.89
113
2.19
50
0.81
111
0.85
121
0.87
122
1.01
123
1.14
114
1.05
111
PWC_ROBbinarytwo views2.69
55
2.01
106
8.36
28
1.33
61
1.90
114
3.23
13
1.18
49
4.33
109
3.25
102
3.11
84
3.41
102
2.71
38
2.18
67
9.15
63
4.48
86
0.56
81
0.31
36
0.61
95
0.36
39
0.69
82
0.74
83
RPtwo views2.79
56
1.20
66
10.36
39
1.54
92
1.93
115
3.44
26
1.97
86
4.05
74
2.34
45
2.66
61
2.79
85
3.98
88
2.58
80
10.21
80
2.61
63
0.95
117
0.47
81
0.71
109
0.57
93
0.71
88
0.68
78
SHDtwo views2.82
57
2.04
108
13.80
53
1.69
103
1.79
111
2.73
6
1.07
42
3.85
44
3.18
97
4.63
110
2.74
83
3.38
66
2.77
85
5.48
35
3.66
81
0.62
88
0.47
81
0.51
72
0.55
88
0.65
73
0.74
83
FADNet-RVC-Resampletwo views2.83
58
2.02
107
21.07
119
1.12
21
0.91
29
3.53
38
1.37
57
4.08
79
2.09
23
1.68
8
1.33
6
2.03
15
1.51
31
9.54
68
1.69
19
0.38
35
0.37
53
0.42
43
0.44
66
0.56
56
0.56
56
StereoDRNettwo views2.84
59
1.49
89
14.48
61
1.33
61
1.57
94
3.44
26
2.06
89
3.69
33
2.36
48
3.26
89
1.84
37
2.79
48
1.80
53
10.88
93
3.16
77
0.41
46
0.33
39
0.39
35
0.36
39
0.59
60
0.58
62
PDISCO_ROBtwo views2.86
60
1.53
90
14.27
60
1.98
120
2.19
122
4.52
111
1.35
56
4.65
121
3.19
98
2.09
34
2.16
63
3.53
73
2.45
76
5.32
32
2.12
48
1.30
130
0.78
118
0.72
112
0.59
96
1.24
121
1.28
117
AF-Nettwo views2.87
61
1.30
76
16.24
76
1.33
61
1.75
107
3.26
16
1.59
72
4.09
80
2.43
54
2.69
62
2.94
89
4.51
103
3.20
94
5.84
42
2.75
68
0.76
106
0.36
48
0.72
112
0.49
75
0.70
83
0.56
56
HSM-Net_RVCpermissivetwo views2.87
61
0.57
5
17.46
90
0.92
4
0.90
28
3.70
51
1.20
51
5.64
131
2.51
64
3.22
88
2.43
74
2.54
32
1.59
36
11.41
97
1.65
15
0.34
21
0.25
20
0.26
6
0.23
11
0.30
8
0.29
11
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
Nwc_Nettwo views2.88
63
1.16
64
18.30
98
1.35
68
1.68
100
3.52
37
1.18
49
3.90
56
2.06
20
2.52
53
2.82
87
3.82
81
2.34
72
6.78
46
2.55
61
0.86
114
0.40
60
0.59
88
0.67
101
0.57
59
0.55
54
MFMNet_retwo views2.91
64
1.91
102
18.25
96
3.06
130
1.89
113
3.13
11
1.67
77
3.52
22
2.97
89
3.14
85
2.87
88
2.26
19
2.41
73
2.42
13
2.00
41
1.10
121
1.05
124
0.94
124
0.90
119
1.28
122
1.46
123
GANettwo views2.93
65
1.05
51
11.29
44
1.64
99
0.96
33
3.91
68
2.52
97
3.72
35
2.41
52
2.95
78
3.00
91
3.21
64
2.03
61
12.42
118
4.72
90
0.51
71
0.41
63
0.46
55
0.39
52
0.60
63
0.50
45
PWCDC_ROBbinarytwo views2.95
66
3.26
118
14.06
56
1.59
95
1.56
91
4.19
95
0.86
19
4.30
106
4.08
115
2.06
33
7.37
114
2.75
41
2.42
74
3.59
19
2.67
66
1.34
132
0.36
48
0.39
35
0.34
31
1.13
113
0.62
66
RTSCtwo views2.95
66
2.51
115
15.49
73
1.54
92
1.30
67
3.51
36
0.86
19
3.89
51
2.48
60
3.64
95
3.12
97
3.04
58
2.30
69
5.70
41
6.67
99
0.55
79
0.35
47
0.37
27
0.39
52
0.70
83
0.60
64
PSMNet_ROBtwo views2.95
66
1.38
81
18.46
101
1.19
32
1.20
57
3.80
62
1.65
76
3.54
23
2.35
46
1.98
25
1.71
23
2.75
41
1.79
52
11.43
99
2.78
69
0.42
48
0.36
48
0.50
70
0.60
97
0.59
60
0.49
43
XQCtwo views2.96
69
2.42
114
14.85
63
1.36
72
1.32
70
3.29
21
1.53
67
4.28
101
2.78
77
3.21
87
2.46
75
3.53
73
2.92
88
6.58
44
4.67
88
0.70
103
0.41
63
0.59
88
0.56
90
0.96
107
0.83
97
SGM-Foresttwo views2.96
69
0.62
10
3.49
17
1.09
15
0.80
14
4.06
83
3.30
111
4.37
110
3.09
93
4.18
105
3.27
99
4.67
107
3.86
107
11.69
107
7.91
108
0.51
71
0.46
79
0.49
68
0.39
52
0.48
37
0.49
43
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
RYNettwo views2.97
71
0.93
38
15.09
69
1.12
21
1.38
75
3.66
49
1.42
63
4.03
70
2.29
41
2.10
35
1.70
18
3.79
80
2.16
66
13.64
125
3.02
75
0.37
29
0.41
63
0.48
62
0.34
31
0.81
95
0.75
85
GwcNetcopylefttwo views3.00
72
1.63
93
19.83
109
1.32
57
2.04
121
3.27
18
1.55
70
3.65
27
2.33
44
2.00
27
2.12
62
3.56
77
1.68
45
9.47
67
2.45
57
0.44
53
0.45
76
0.46
55
0.38
49
0.60
63
0.76
86
AANet_RVCtwo views3.03
73
1.72
98
12.78
49
1.17
28
1.12
49
3.79
61
1.10
43
3.86
46
3.13
94
2.62
59
4.26
109
3.96
87
1.41
16
11.43
99
5.23
97
0.94
116
0.47
81
0.34
24
0.23
11
0.40
22
0.55
54
DRN-Testtwo views3.03
73
0.99
46
18.18
95
1.35
68
1.60
96
3.59
43
1.48
64
4.13
84
2.45
57
3.47
94
1.92
44
2.58
35
2.53
78
10.88
93
2.80
70
0.39
40
0.36
48
0.44
49
0.41
61
0.55
53
0.57
59
Anonymous Stereotwo views3.09
75
3.14
117
16.69
84
1.52
89
1.25
63
3.33
22
2.95
104
4.02
69
2.42
53
2.35
48
2.04
51
2.57
34
1.27
12
11.88
112
2.92
72
0.49
65
0.49
87
0.57
85
0.56
90
0.70
83
0.65
73
stereogantwo views3.09
75
0.93
38
16.43
79
1.37
75
1.66
99
3.70
51
1.67
77
3.87
48
3.15
96
3.04
80
2.70
82
5.04
113
2.79
86
9.96
77
1.78
29
0.63
89
0.49
87
0.69
105
0.47
70
0.83
98
0.68
78
PA-Nettwo views3.09
75
1.24
73
18.84
102
1.38
76
1.19
55
3.42
25
1.75
79
3.89
51
3.27
106
1.88
18
2.10
58
3.03
57
1.88
59
11.60
101
2.88
71
0.40
41
0.54
90
0.44
49
0.62
99
0.52
43
1.02
109
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
RGCtwo views3.10
78
1.45
87
14.90
65
1.46
81
1.79
111
3.33
22
1.39
59
4.04
72
2.26
37
2.77
67
3.26
98
4.00
89
2.71
84
11.63
104
2.62
65
0.82
112
0.46
79
0.79
117
0.79
114
0.71
88
0.78
91
PVDtwo views3.12
79
1.97
105
12.80
50
1.79
109
1.63
97
3.83
64
1.34
55
3.48
19
3.88
114
5.77
114
3.56
103
3.45
68
3.29
96
5.58
38
5.30
98
0.71
104
0.69
114
0.74
115
0.77
111
0.79
94
1.13
113
Abc-Nettwo views3.16
80
1.28
74
18.31
99
1.25
44
1.75
107
4.06
83
1.40
61
3.89
51
2.22
31
2.72
64
2.55
77
3.49
69
2.65
82
9.91
74
2.99
73
1.29
128
0.65
107
0.64
101
0.86
117
0.65
73
0.71
80
Xing Li, Yangyu Fan, Guoyun Lv, and Haoyue Ma: Area-based Correlation and Non-local Attention Network for Stereo Matching. The Visual Computer
NCC-stereotwo views3.16
80
1.28
74
18.31
99
1.25
44
1.75
107
4.06
83
1.40
61
3.89
51
2.22
31
2.72
64
2.55
77
3.49
69
2.65
82
9.91
74
2.99
73
1.29
128
0.65
107
0.64
101
0.86
117
0.65
73
0.71
80
GANetREF_RVCpermissivetwo views3.20
82
1.93
103
18.84
102
1.51
88
0.80
14
4.12
91
1.82
83
4.16
89
2.91
88
1.87
17
1.74
28
2.48
28
1.84
56
13.18
122
1.74
28
0.80
109
0.57
94
0.79
117
0.74
107
1.20
120
0.86
98
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
FBW_ROBtwo views3.24
83
1.02
48
9.21
34
1.49
85
1.07
45
3.78
58
1.48
64
4.31
107
3.08
92
2.46
50
2.32
72
4.20
96
2.08
63
15.19
130
7.16
107
0.84
113
0.62
104
1.33
131
0.92
120
0.99
108
1.31
118
aanetorigintwo views3.30
84
1.66
97
17.52
92
1.18
31
1.58
95
2.57
4
3.80
119
2.65
4
2.03
17
6.12
118
4.30
110
2.82
49
5.70
117
5.47
34
4.67
88
0.50
68
0.44
72
0.43
46
0.47
70
1.17
117
0.92
105
ADCReftwo views3.30
84
1.23
72
14.90
65
1.17
28
1.41
79
3.95
73
1.64
75
3.88
50
2.36
48
3.05
82
2.21
69
2.40
25
2.81
87
5.04
28
16.63
123
0.46
59
0.34
46
0.63
98
0.62
99
0.54
50
0.67
76
CBMVpermissivetwo views3.32
86
0.96
45
21.16
121
1.32
57
0.83
20
3.81
63
3.91
122
4.47
116
2.53
65
3.34
92
2.24
70
4.45
101
2.58
80
10.07
79
1.96
39
0.45
58
0.43
69
0.48
62
0.38
49
0.52
43
0.53
50
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
FAT-Stereotwo views3.34
87
0.81
28
20.05
110
1.25
44
1.20
57
3.44
26
1.99
88
4.23
97
2.67
71
3.14
85
3.82
107
3.51
71
3.32
98
11.68
106
1.94
37
0.64
92
0.67
110
0.48
62
0.52
82
0.61
67
0.81
93
edge stereotwo views3.37
88
1.01
47
21.99
127
1.28
51
1.37
73
2.67
5
1.39
59
3.98
65
2.87
85
2.93
76
3.27
99
4.35
99
2.43
75
11.62
103
2.39
55
0.60
84
0.63
105
0.79
117
0.55
88
0.68
80
0.67
76
DPSNettwo views3.43
89
1.37
80
12.19
46
1.39
77
1.24
61
4.15
94
2.06
89
4.51
117
3.34
107
2.69
62
1.96
45
3.52
72
5.56
116
10.68
89
9.68
113
0.93
115
0.73
116
0.40
38
0.37
47
1.06
110
0.86
98
S-Stereotwo views3.48
90
0.91
36
22.06
128
1.36
72
1.50
83
3.55
41
2.53
98
4.21
94
2.81
82
2.37
49
3.81
106
3.75
79
1.84
56
11.90
114
2.51
60
0.63
89
0.71
115
0.48
62
0.76
109
0.67
78
1.19
115
SAMSARAtwo views3.55
91
2.16
111
14.18
57
1.86
115
1.49
82
4.62
112
2.65
101
3.58
26
2.80
80
7.05
121
3.05
94
5.16
115
4.07
108
9.18
64
4.80
91
0.64
92
0.67
110
0.59
88
0.68
103
0.89
100
0.91
104
CBMV_ROBtwo views3.62
92
0.70
17
21.42
125
1.21
36
0.89
26
4.43
110
1.76
80
4.39
112
2.79
78
5.88
115
3.38
101
4.18
94
5.12
114
10.75
90
2.32
54
0.56
81
0.55
92
0.53
79
0.51
79
0.55
53
0.46
38
DeepPrunerFtwo views3.70
93
1.72
98
14.87
64
1.33
61
1.74
106
3.70
51
1.55
70
4.20
92
14.00
133
2.02
28
1.77
30
2.74
39
2.30
69
10.88
93
6.88
104
0.68
100
0.60
101
0.78
116
0.78
113
0.81
95
0.71
80
ADCP+two views3.74
94
1.96
104
16.90
88
1.09
15
1.55
88
4.35
105
3.65
118
3.93
59
2.98
90
2.22
40
1.66
16
3.95
85
4.83
113
3.41
17
18.87
129
0.38
35
0.41
63
0.63
98
0.53
86
0.64
72
0.81
93
MFN_U_SF_RVCtwo views3.81
95
10.78
136
15.01
68
1.84
112
1.39
76
3.73
55
1.63
73
4.05
74
3.26
103
2.91
74
2.38
73
4.36
100
3.65
103
8.43
54
4.91
93
1.11
122
1.04
123
0.99
125
1.31
129
1.58
128
1.75
129
ADCPNettwo views3.89
96
1.63
93
20.25
112
1.28
51
1.27
66
6.43
125
2.65
101
3.93
59
2.70
73
2.78
68
2.20
67
4.28
97
6.16
120
5.56
37
10.27
116
0.65
95
1.18
127
0.71
109
1.70
132
0.84
99
1.37
120
RTStwo views3.90
97
7.04
128
17.94
93
1.92
117
2.19
122
5.00
117
2.96
107
4.29
102
3.19
98
4.05
102
3.08
95
4.94
110
3.16
92
7.64
49
6.72
100
0.66
96
0.44
72
0.51
72
0.47
70
0.95
103
0.76
86
RTSAtwo views3.90
97
7.04
128
17.94
93
1.92
117
2.19
122
5.00
117
2.96
107
4.29
102
3.19
98
4.05
102
3.08
95
4.94
110
3.16
92
7.64
49
6.72
100
0.66
96
0.44
72
0.51
72
0.47
70
0.95
103
0.76
86
MFN_U_SF_DS_RVCtwo views3.94
99
10.88
137
12.30
47
1.26
47
1.55
88
5.19
119
3.33
112
4.42
113
2.82
84
2.58
56
2.62
80
5.00
112
2.11
64
14.03
128
2.73
67
1.28
127
2.34
134
0.67
104
1.83
133
0.81
95
1.03
110
ADCMidtwo views3.96
100
3.27
119
15.48
72
1.26
47
1.40
77
4.66
113
2.28
94
4.06
76
3.54
109
3.94
101
3.02
92
3.85
82
3.67
104
4.03
22
18.65
127
0.68
100
0.61
103
1.09
129
1.18
127
1.38
125
1.08
112
MADNet+two views3.97
101
9.54
133
13.83
54
1.87
116
1.69
101
4.19
95
2.87
103
4.80
125
4.10
116
2.51
52
2.67
81
4.55
105
3.41
99
8.72
56
6.92
105
1.25
126
1.27
130
1.04
126
0.93
121
1.39
126
1.85
130
SPS-STEREOcopylefttwo views3.97
101
1.44
86
20.85
117
1.72
106
1.71
103
3.24
14
2.30
95
3.55
24
2.68
72
4.79
111
3.66
104
5.44
116
2.97
89
9.11
62
9.35
112
1.15
123
1.10
125
1.06
128
1.02
125
1.16
116
1.18
114
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
NaN_ROBtwo views4.06
103
1.13
61
16.64
82
1.12
21
1.25
63
3.50
35
3.05
109
4.29
102
3.26
103
3.64
95
2.56
79
2.97
55
4.79
112
13.33
123
16.91
124
0.37
29
0.41
63
0.38
30
0.39
52
0.48
37
0.81
93
STTStereo_v2two views4.11
104
1.21
67
19.47
107
2.31
126
2.94
129
4.74
114
2.95
104
3.33
16
2.14
25
4.44
107
9.43
123
6.93
119
3.84
105
9.01
60
4.92
94
0.79
107
0.80
119
0.70
107
0.50
76
0.95
103
0.87
101
G-Nettwo views4.11
104
1.21
67
19.47
107
2.31
126
2.94
129
4.74
114
2.95
104
3.33
16
2.14
25
4.44
107
9.43
123
6.93
119
3.84
105
9.01
60
4.92
94
0.79
107
0.80
119
0.70
107
0.50
76
0.95
103
0.87
101
AnyNet_C32two views4.17
106
5.75
126
12.62
48
1.60
96
1.54
84
5.51
120
4.09
123
3.90
56
3.23
101
4.22
106
3.03
93
3.10
59
3.41
99
8.01
51
18.85
128
0.53
78
0.58
97
0.53
79
0.74
107
1.19
119
0.97
107
ADCStwo views4.22
107
5.33
125
16.32
77
1.64
99
1.31
68
3.91
68
3.64
117
4.21
94
2.89
87
5.11
112
2.81
86
4.46
102
3.06
91
5.21
31
19.50
130
0.63
89
0.64
106
0.69
105
0.69
105
1.12
112
1.26
116
ADCLtwo views4.35
108
1.63
93
14.21
58
1.29
54
1.26
65
5.56
121
11.16
130
3.97
64
3.37
108
6.07
116
4.67
111
4.30
98
5.72
118
5.08
29
14.51
120
0.50
68
0.48
85
0.89
123
0.77
111
0.78
92
0.76
86
AnyNet_C01two views4.37
109
7.80
130
16.40
78
1.98
120
1.40
77
6.37
124
6.47
125
4.09
80
5.03
122
4.06
104
3.78
105
4.16
93
3.31
97
6.62
45
10.00
114
0.61
86
0.67
110
0.63
98
0.67
101
2.01
130
1.31
118
MSC_U_SF_DS_RVCtwo views4.48
110
10.62
135
15.18
70
1.55
94
1.45
80
6.16
123
3.54
116
5.41
130
3.68
112
2.59
57
2.74
83
4.81
108
2.20
68
13.79
126
4.40
85
2.04
133
3.36
135
1.05
127
1.20
128
2.09
131
1.72
128
SGM_RVCbinarytwo views4.57
111
0.72
19
17.46
90
1.84
112
0.58
6
4.23
102
3.45
113
4.66
122
3.66
111
7.99
122
8.82
122
7.56
122
8.74
126
12.30
116
6.84
103
0.43
49
0.38
55
0.39
35
0.35
34
0.51
41
0.48
40
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
DispFullNettwo views4.64
112
3.48
120
23.95
131
3.99
131
3.03
131
3.35
24
1.06
41
4.18
90
2.80
80
6.33
119
2.94
89
4.04
90
7.81
123
3.92
21
5.16
96
2.75
134
1.61
133
3.10
135
2.68
134
3.32
132
3.31
134
CSANtwo views4.71
113
1.39
83
20.23
111
1.19
32
0.87
23
3.65
48
3.50
115
4.57
119
3.87
113
3.89
100
7.82
117
3.54
76
6.16
120
23.47
138
6.82
102
0.52
76
0.45
76
0.49
68
0.52
82
0.63
68
0.66
75
SuperBtwo views4.79
114
4.41
123
21.08
120
1.45
80
1.08
46
3.91
68
3.08
110
4.18
90
3.01
91
2.50
51
2.10
58
2.75
41
3.49
102
8.80
58
19.60
131
0.44
53
0.33
39
0.53
79
0.39
52
12.01
138
0.64
70
NVStereoNet_ROBtwo views4.91
115
1.40
84
16.46
80
1.47
83
1.72
105
3.46
30
1.96
85
3.89
51
3.61
110
2.90
73
24.61
139
4.81
108
4.49
110
13.57
124
4.87
92
1.33
131
1.29
131
1.70
133
1.68
131
1.41
127
1.66
125
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
WCMA_ROBtwo views5.12
116
0.90
34
21.33
123
1.50
86
1.33
71
4.09
86
3.49
114
4.03
70
4.59
118
6.57
120
10.95
125
12.92
128
8.41
125
9.86
73
8.78
111
0.75
105
0.58
97
0.51
72
0.50
76
0.68
80
0.57
59
MeshStereopermissivetwo views5.35
117
1.32
77
21.34
124
1.34
65
1.15
52
4.40
108
3.90
121
5.18
126
4.55
117
10.14
130
11.82
128
11.47
126
5.84
119
13.84
127
7.03
106
0.61
86
0.58
97
0.60
92
0.58
94
0.72
90
0.59
63
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
pmcnntwo views5.45
118
1.22
70
17.40
89
1.60
96
2.41
126
3.48
34
2.12
92
4.52
118
3.26
103
3.67
97
13.15
129
22.10
134
3.46
101
11.92
115
17.16
125
0.34
21
0.21
7
0.23
2
0.19
4
0.36
15
0.28
10
MDST_ROBtwo views5.53
119
0.59
9
23.11
129
2.55
128
2.69
127
21.20
135
1.98
87
4.77
124
2.65
69
9.36
126
4.25
108
4.14
92
4.45
109
22.57
134
3.68
82
0.48
64
0.33
39
0.47
57
0.44
66
0.44
29
0.44
34
MSMD_ROBtwo views5.61
120
1.08
55
16.68
83
1.21
36
0.83
20
10.14
130
2.07
91
3.93
59
6.22
127
10.37
131
8.03
119
22.22
135
9.45
127
11.66
105
4.52
87
0.66
96
0.58
97
0.66
103
0.68
103
0.65
73
0.63
69
EDNetEfficienttwo views5.61
120
2.14
110
16.75
86
1.67
101
3.64
132
4.12
91
7.51
126
3.05
9
5.63
126
11.57
134
4.83
113
3.95
85
12.72
130
8.44
55
20.78
133
0.41
46
0.43
69
0.47
57
1.01
123
1.66
129
1.39
122
SANettwo views5.65
122
1.72
98
20.94
118
1.26
47
0.88
24
7.20
127
3.81
120
4.60
120
11.23
130
4.56
109
8.78
121
7.49
121
6.95
122
11.85
111
18.11
126
0.49
65
0.44
72
0.50
70
0.42
64
0.90
102
0.82
96
ELAS_RVCcopylefttwo views5.72
123
1.64
96
20.40
115
2.09
124
2.01
118
4.95
116
11.47
131
4.66
122
6.31
128
8.09
123
8.01
118
8.06
125
11.65
129
8.36
53
10.22
115
0.99
120
1.17
126
0.81
120
0.76
109
1.14
114
1.58
124
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
PWCKtwo views5.85
124
9.95
134
15.22
71
2.90
129
1.56
91
5.66
122
7.69
127
5.27
128
4.74
120
3.88
99
7.51
115
8.04
123
5.28
115
15.18
129
8.40
109
3.08
135
1.51
132
2.24
134
1.37
130
5.55
134
1.96
131
ELAScopylefttwo views5.99
125
1.62
92
20.29
113
1.96
119
2.01
118
15.09
134
8.38
129
5.23
127
5.54
124
9.02
125
8.61
120
6.74
118
9.75
128
7.48
48
11.40
118
0.96
118
1.18
127
0.72
112
0.93
121
1.30
123
1.67
126
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
FCDSN-DCtwo views6.04
126
0.84
33
19.25
104
1.24
42
1.09
47
4.38
106
2.62
100
4.38
111
4.71
119
9.59
129
11.22
126
19.19
131
17.40
133
12.30
116
8.55
110
0.69
102
0.57
94
0.60
92
0.60
97
0.70
83
0.78
91
Dominik Hirner, Friedrich Fraundorfer: FCDSN-DC: An accurate but lightweight end-to-end trainable neural network for stereo estimation with depth completion.
psmorigintwo views6.59
127
1.15
63
88.11
141
1.52
89
0.97
37
2.88
7
1.20
51
3.21
12
2.04
18
3.73
98
2.46
75
5.15
114
2.55
79
10.60
87
2.47
58
0.55
79
0.68
113
0.48
62
0.51
79
0.70
83
0.88
103
FC-DCNNcopylefttwo views6.64
128
0.77
25
19.26
105
1.68
102
1.64
98
8.27
129
2.35
96
6.35
132
9.23
129
10.85
132
11.27
127
19.89
132
13.33
131
12.51
121
11.79
119
0.66
96
0.57
94
0.58
87
0.56
90
0.63
68
0.61
65
MADNet++two views7.57
129
5.80
127
13.28
52
6.38
134
5.26
134
7.18
126
7.77
128
5.32
129
5.62
125
6.11
117
7.78
116
6.02
117
8.01
124
11.74
109
16.36
122
9.36
138
4.95
136
8.57
138
5.66
136
4.83
133
5.34
135
SGM+DAISYtwo views8.36
130
3.82
122
24.83
132
2.02
122
1.75
107
14.80
133
17.22
133
4.27
100
11.79
132
13.88
135
14.38
132
11.74
127
13.83
132
10.56
86
15.12
121
1.21
124
1.19
129
1.14
130
1.13
126
1.17
117
1.37
120
MANEtwo views11.39
131
1.38
81
23.16
130
4.82
133
2.79
128
22.02
136
14.51
132
25.81
137
26.28
138
10.93
133
13.17
130
19.90
133
18.03
134
15.69
131
20.41
132
0.96
118
0.74
117
1.42
132
3.24
135
0.74
91
1.71
127
SGM-ForestMtwo views11.55
132
1.16
64
21.62
126
2.14
125
0.81
17
22.25
137
19.08
134
16.50
136
11.34
131
15.94
138
14.44
133
27.12
138
22.61
139
24.76
139
28.39
138
0.50
68
0.48
85
0.47
57
0.44
66
0.55
53
0.53
50
EDNetEfficientorigintwo views12.35
133
2.51
115
179.47
144
1.19
32
0.96
33
2.91
8
1.81
82
3.00
8
5.35
123
8.91
124
4.79
112
4.18
94
4.64
111
9.98
78
11.30
117
0.37
29
0.38
55
0.60
92
0.85
116
1.32
124
2.43
133
LSMtwo views12.71
134
3.52
121
20.79
116
55.63
142
84.76
142
3.63
47
6.30
124
3.73
36
4.93
121
5.55
113
13.92
131
3.53
73
3.25
95
9.72
69
4.12
84
0.64
92
0.85
121
0.71
109
0.81
115
1.03
109
26.84
140
DGTPSM_ROBtwo views15.31
135
8.85
131
28.14
135
9.90
136
18.89
135
12.44
131
33.21
137
13.48
134
17.86
136
9.50
127
22.89
137
13.78
129
19.67
135
21.26
132
23.68
134
5.08
136
8.84
137
5.62
136
11.50
137
6.72
135
14.87
136
DPSMNet_ROBtwo views15.52
136
8.89
132
31.44
136
10.07
137
18.91
136
12.44
131
33.22
138
13.51
135
17.94
137
9.58
128
22.91
138
13.79
130
19.67
135
21.43
133
23.73
135
5.17
137
8.84
137
5.63
137
11.52
140
6.79
136
14.90
137
LE_ROBtwo views18.58
137
1.21
67
48.38
138
8.60
135
23.00
139
7.32
128
19.74
135
9.94
133
38.69
140
43.53
140
21.91
136
28.40
139
40.92
141
23.18
135
54.71
141
0.34
21
0.33
39
0.32
16
0.36
39
0.42
26
0.40
28
DPSMtwo views20.78
138
20.03
138
26.57
133
23.62
138
22.46
137
31.10
138
39.97
140
33.52
139
17.41
134
15.25
136
15.80
134
24.06
136
20.71
137
23.30
136
27.62
136
10.78
139
9.41
139
11.69
139
11.50
137
14.49
139
16.36
138
DPSM_ROBtwo views20.78
138
20.03
138
26.57
133
23.62
138
22.46
137
31.10
138
39.97
140
33.52
139
17.41
134
15.25
136
15.80
134
24.06
136
20.71
137
23.30
136
27.62
136
10.78
139
9.41
139
11.69
139
11.50
137
14.49
139
16.36
138
AVERAGE_ROBtwo views44.46
140
45.91
141
46.59
137
40.59
141
38.66
140
32.99
140
30.52
136
46.19
141
43.26
141
49.59
141
48.97
141
44.62
140
39.89
140
45.04
140
43.94
139
48.74
142
50.69
142
50.80
142
49.79
142
45.54
142
46.97
141
MEDIAN_ROBtwo views47.37
141
49.08
142
49.92
139
40.32
140
39.97
141
34.95
141
33.43
139
49.17
142
46.39
142
53.06
142
52.65
142
47.65
141
42.23
143
48.35
142
47.29
140
52.10
143
54.04
143
54.28
143
53.48
143
48.76
143
50.33
142
LSM0two views49.58
142
40.04
140
53.62
140
97.15
143
128.68
144
62.75
142
79.89
143
67.63
143
35.11
139
33.02
139
32.61
140
48.43
142
41.60
142
46.60
141
55.43
142
21.63
141
18.99
141
23.62
141
23.13
141
29.47
141
52.26
143
BEATNet-Init1two views49.72
143
64.32
143
97.77
142
4.68
132
4.28
133
84.75
143
60.42
142
73.38
144
68.59
143
88.29
143
88.51
144
91.94
143
97.99
144
85.26
144
70.96
143
0.80
109
0.65
107
0.85
121
0.70
106
7.96
137
2.33
132
DPSimNet_ROBtwo views102.30
144
119.72
145
143.59
143
109.42
144
94.10
143
127.22
144
127.15
144
26.62
138
102.36
144
149.58
144
83.99
143
139.34
144
137.54
145
64.62
143
82.36
144
105.18
144
66.15
144
120.15
144
69.61
144
99.58
144
77.72
144
MSMDNettwo views1.11
7
ASD4two views113.98
144