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
DPM-Stereotwo views1.97
1
0.64
18
2.95
10
0.17
1
0.10
1
4.83
31
0.13
2
8.60
11
4.06
4
6.42
22
4.92
4
0.44
1
0.72
1
3.57
4
1.80
5
0.00
1
0.01
31
0.00
1
0.00
1
0.05
16
0.04
20
PMTNettwo views1.99
2
0.32
1
2.21
3
0.39
2
0.23
6
5.08
33
0.49
12
5.84
1
8.22
34
3.07
1
3.29
1
0.73
2
0.75
2
8.18
34
0.94
3
0.00
1
0.00
1
0.00
1
0.00
1
0.03
10
0.00
1
R-Stereo Traintwo views2.44
3
0.32
1
1.93
1
0.94
5
0.16
4
3.67
13
0.61
19
6.37
3
3.08
1
9.14
43
17.44
72
1.80
3
0.77
3
1.76
1
0.70
1
0.00
1
0.01
31
0.00
1
0.00
1
0.01
1
0.03
11
R-Stereotwo views2.44
3
0.32
1
1.93
1
0.94
5
0.16
4
3.67
13
0.61
19
6.37
3
3.08
1
9.14
43
17.44
72
1.80
3
0.77
3
1.76
1
0.70
1
0.00
1
0.01
31
0.00
1
0.00
1
0.01
1
0.03
11
DN-CSS_ROBtwo views2.69
5
1.40
62
5.34
39
2.31
56
0.75
23
3.14
11
0.06
1
6.11
2
3.87
3
5.34
12
12.18
46
2.34
5
1.22
5
7.84
24
1.48
4
0.03
39
0.00
1
0.00
1
0.00
1
0.35
69
0.03
11
HITNettwo views2.79
6
0.77
21
4.02
24
2.03
42
0.11
3
5.58
37
0.59
18
9.24
16
5.15
8
6.42
22
7.26
14
3.66
6
2.92
18
4.07
5
3.87
32
0.00
1
0.00
1
0.00
1
0.00
1
0.06
20
0.02
3
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
DMCAtwo views2.82
7
0.64
18
4.45
30
1.61
24
0.89
29
3.86
17
0.57
17
9.18
14
5.10
7
6.24
18
6.49
8
7.17
25
2.13
10
3.31
3
4.73
38
0.00
1
0.01
31
0.00
1
0.00
1
0.04
13
0.02
3
ccstwo views3.04
8
0.39
7
3.08
12
1.78
30
0.52
16
2.04
1
0.50
13
13.09
60
13.71
66
3.54
4
5.36
7
5.50
12
2.45
12
4.81
7
2.88
13
0.09
55
0.08
68
0.12
83
0.10
77
0.20
47
0.50
79
AdaStereotwo views3.09
9
0.58
13
3.04
11
2.84
70
0.48
15
4.08
20
1.29
37
12.16
52
7.77
29
6.03
16
9.62
25
5.79
14
1.53
7
4.56
6
1.93
7
0.00
1
0.00
1
0.00
1
0.00
1
0.02
4
0.02
3
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.
DMCA-RVCcopylefttwo views3.17
10
1.26
55
6.39
57
2.10
45
1.51
46
3.02
10
0.56
16
7.84
7
5.06
6
9.23
46
5.19
5
8.46
33
2.72
16
6.31
12
3.22
20
0.06
51
0.11
74
0.09
80
0.02
49
0.16
41
0.08
27
BEATNet_4xtwo views3.24
11
1.27
56
5.89
48
1.56
21
0.10
1
5.26
34
1.07
32
10.08
23
5.50
9
6.89
27
7.73
16
4.53
9
4.13
30
5.05
8
5.27
41
0.04
45
0.05
55
0.00
1
0.00
1
0.23
60
0.23
55
GwcNet-RSSMtwo views3.27
12
0.60
14
3.38
16
2.94
76
2.00
60
2.61
5
0.48
10
10.93
40
6.11
11
5.47
13
10.20
29
6.57
18
5.44
50
6.25
11
2.23
9
0.00
1
0.00
1
0.08
77
0.00
1
0.17
43
0.03
11
NOSS_ROBtwo views3.30
13
0.46
8
2.62
4
2.08
43
1.01
36
5.60
38
0.74
28
10.37
30
11.48
56
5.15
10
8.43
21
5.67
13
1.73
8
7.97
26
2.34
10
0.02
29
0.06
61
0.00
1
0.00
1
0.07
21
0.14
46
CFNet_RVCtwo views3.31
14
0.94
34
2.69
5
1.50
18
2.38
67
2.81
6
0.68
24
8.35
8
7.43
23
4.45
6
9.94
26
10.20
43
4.60
35
6.49
14
3.41
24
0.00
1
0.00
1
0.03
64
0.00
1
0.22
55
0.03
11
CFNet-ftpermissivetwo views3.31
14
0.94
34
2.69
5
1.50
18
2.38
67
2.81
6
0.68
24
8.35
8
7.43
23
4.45
6
9.94
26
10.20
43
4.60
35
6.50
15
3.41
24
0.00
1
0.00
1
0.03
64
0.00
1
0.22
55
0.03
11
PSMNet-RSSMtwo views3.33
16
0.77
21
3.22
15
2.12
46
1.90
59
2.33
3
0.71
27
11.17
41
5.83
10
5.90
14
11.43
41
7.04
21
4.83
44
5.96
10
3.17
18
0.00
1
0.00
1
0.02
56
0.00
1
0.08
23
0.02
3
cf-rtwo views3.42
17
0.81
24
3.97
23
2.41
58
2.78
76
2.85
8
0.64
23
8.44
10
6.32
12
6.25
19
11.57
42
8.97
37
3.54
22
5.93
9
3.83
31
0.00
1
0.00
1
0.00
1
0.00
1
0.09
26
0.07
24
MLCVtwo views3.44
18
0.88
27
5.60
43
1.39
14
0.25
7
4.36
24
0.33
6
7.25
5
7.28
20
9.17
45
12.24
48
5.09
10
2.47
13
9.15
46
3.23
21
0.00
1
0.00
1
0.00
1
0.00
1
0.10
27
0.02
3
DeepPruner_ROBtwo views3.52
19
1.14
49
4.06
25
1.12
8
1.65
50
3.65
12
0.83
30
13.96
68
4.47
5
7.80
32
10.84
34
7.05
23
2.16
11
8.14
32
3.08
17
0.07
53
0.03
47
0.00
1
0.01
45
0.32
65
0.06
23
STTStereotwo views3.60
20
0.93
33
6.34
55
2.71
68
2.23
65
3.68
15
0.63
22
9.42
17
6.73
15
9.87
52
6.97
12
8.84
36
3.65
23
6.85
16
3.04
15
0.00
1
0.02
42
0.01
45
0.00
1
0.02
4
0.02
3
ccs_robtwo views3.63
21
1.12
48
4.42
28
2.52
61
0.91
32
5.50
36
0.21
4
10.11
26
9.11
40
6.55
25
11.28
38
8.32
32
2.55
14
7.66
21
2.01
8
0.00
1
0.00
1
0.00
1
0.00
1
0.20
47
0.08
27
GwcNetcopylefttwo views3.66
22
0.65
20
2.79
8
0.89
4
0.72
22
4.25
22
0.55
15
9.11
13
7.52
25
7.14
29
10.19
28
12.49
56
4.70
39
7.44
20
4.41
35
0.03
39
0.00
1
0.06
75
0.00
1
0.20
47
0.04
20
iResNettwo views3.68
23
0.91
30
7.94
72
2.97
77
0.34
9
4.44
28
0.48
10
7.70
6
9.74
44
7.72
31
12.74
51
4.03
7
2.87
17
8.05
28
3.37
23
0.02
29
0.01
31
0.00
1
0.00
1
0.10
27
0.09
31
CFNettwo views3.72
24
1.10
46
5.03
34
2.49
60
1.59
47
4.90
32
0.22
5
11.38
42
9.88
46
4.80
8
11.25
37
6.44
17
3.68
24
8.33
35
3.00
14
0.00
1
0.00
1
0.00
1
0.00
1
0.22
55
0.07
24
GA-Net-RSSMtwo views3.77
25
1.05
40
5.06
35
2.21
51
1.87
57
2.36
4
1.52
40
8.98
12
7.96
30
5.91
15
14.59
60
7.94
30
3.50
21
6.43
13
5.70
43
0.00
1
0.00
1
0.04
70
0.00
1
0.20
47
0.08
27
FADNet-RVC-Resampletwo views3.79
26
1.62
73
12.06
85
1.43
16
0.66
18
5.94
40
2.41
46
10.18
28
8.58
38
6.28
20
4.22
3
5.33
11
4.80
43
7.71
22
3.19
19
0.17
67
0.21
88
0.17
89
0.12
79
0.41
77
0.29
67
NLCA_NET_v2_RVCtwo views3.84
27
1.06
41
5.23
37
2.72
69
3.27
81
4.36
24
0.61
19
10.71
35
7.56
26
8.75
39
7.89
17
9.86
42
3.90
27
7.15
18
3.44
26
0.14
62
0.02
42
0.02
56
0.03
58
0.04
13
0.03
11
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
CC-Net-ROBtwo views3.84
27
1.07
42
5.23
37
2.65
65
2.96
79
4.22
21
0.69
26
10.43
31
7.72
27
8.78
40
8.29
20
9.61
40
4.02
29
7.16
19
3.65
29
0.13
61
0.03
47
0.02
56
0.03
58
0.05
16
0.03
11
FADNet_RVCtwo views3.91
29
1.67
76
12.95
91
0.96
7
0.75
23
5.71
39
0.54
14
10.83
38
6.60
14
3.46
2
8.09
18
4.10
8
3.40
20
9.43
49
6.33
46
0.36
83
0.44
100
0.17
89
0.46
102
0.91
93
0.95
96
FADNet-RVCtwo views3.98
30
1.84
82
12.48
88
1.69
28
0.44
13
4.33
23
1.31
38
11.84
46
7.15
18
3.53
3
3.50
2
10.63
47
4.43
34
9.12
45
6.25
45
0.03
39
0.10
72
0.00
1
0.03
58
0.60
83
0.25
61
HSMtwo views4.00
31
0.79
23
3.16
14
1.59
23
2.17
63
6.77
46
1.11
33
12.28
53
6.35
13
6.75
26
8.11
19
13.90
62
5.37
49
8.85
43
2.71
12
0.00
1
0.00
1
0.00
1
0.00
1
0.02
4
0.02
3
TDLMtwo views4.11
32
1.11
47
3.54
18
1.62
25
1.04
37
3.91
18
7.41
90
10.60
34
10.67
50
6.38
21
12.59
50
5.95
15
4.77
41
8.79
42
3.04
15
0.58
95
0.00
1
0.01
45
0.00
1
0.19
46
0.12
41
CBMV_ROBtwo views4.14
33
0.52
10
3.14
13
1.30
11
0.77
26
6.92
47
1.97
45
10.11
26
9.58
42
8.92
42
14.20
59
7.12
24
5.90
53
8.65
39
3.50
28
0.01
24
0.05
55
0.00
1
0.00
1
0.04
13
0.09
31
CVANet_RVCtwo views4.16
34
1.16
50
3.60
19
1.94
40
1.46
44
3.92
19
4.68
72
10.89
39
8.34
36
7.58
30
10.84
34
10.27
45
6.62
57
8.56
38
2.69
11
0.39
85
0.00
1
0.00
1
0.01
45
0.21
54
0.09
31
HSM-Net_RVCpermissivetwo views4.20
35
0.32
1
2.76
7
0.63
3
0.69
20
6.95
48
1.69
43
11.96
47
8.36
37
8.83
41
12.17
45
15.18
70
4.21
32
6.91
17
3.30
22
0.02
29
0.02
42
0.00
1
0.00
1
0.01
1
0.01
2
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
DSFCAtwo views4.21
36
0.50
9
5.45
41
1.34
12
1.68
52
7.04
49
4.51
70
10.73
36
7.00
16
10.78
58
6.80
9
8.48
34
4.63
37
9.91
53
5.12
40
0.02
29
0.06
61
0.02
56
0.03
58
0.08
23
0.03
11
iResNet_ROBtwo views4.23
37
1.02
38
4.90
33
2.18
48
0.93
34
2.92
9
0.37
8
15.10
78
16.91
80
7.89
34
10.51
32
7.03
20
3.07
19
8.16
33
3.46
27
0.01
24
0.00
1
0.00
1
0.00
1
0.10
27
0.02
3
FADNettwo views4.23
37
1.65
75
11.75
84
1.64
27
0.80
28
4.80
30
0.77
29
13.76
67
11.65
58
3.97
5
5.24
6
9.62
41
5.14
46
8.40
36
3.78
30
0.21
71
0.04
51
0.07
76
0.05
69
1.14
98
0.10
38
iResNetv2_ROBtwo views4.28
39
1.43
63
7.17
67
2.91
71
1.26
41
4.36
24
1.62
41
13.64
66
10.25
49
9.83
51
11.41
39
7.68
27
4.00
28
7.75
23
1.85
6
0.00
1
0.00
1
0.00
1
0.00
1
0.37
71
0.09
31
StereoDRNet-Refinedtwo views4.46
40
0.62
17
3.80
22
1.92
37
0.40
11
9.35
59
0.15
3
10.02
21
8.83
39
12.69
70
11.62
43
9.34
38
3.87
26
8.06
29
8.02
59
0.00
1
0.00
1
0.01
45
0.05
69
0.20
47
0.26
64
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
DLCB_ROBtwo views4.51
41
0.91
30
3.78
21
2.19
49
1.07
38
6.28
41
3.09
50
9.78
20
7.72
27
10.65
56
12.97
52
13.91
63
3.71
25
8.72
40
5.30
42
0.00
1
0.00
1
0.00
1
0.00
1
0.03
10
0.10
38
NVstereo2Dtwo views4.51
41
0.82
25
6.86
65
3.28
82
3.38
85
8.16
54
3.13
51
10.51
32
15.15
71
4.90
9
6.89
11
7.87
28
4.78
42
9.88
52
3.91
33
0.01
24
0.00
1
0.00
1
0.06
71
0.02
4
0.58
84
RASNettwo views4.52
43
0.61
16
4.42
28
3.42
85
4.68
103
4.58
29
0.99
31
9.54
19
8.01
31
5.28
11
11.42
40
10.34
46
8.88
69
9.28
47
8.68
67
0.15
64
0.00
1
0.00
1
0.00
1
0.03
10
0.04
20
SGM-Foresttwo views4.96
44
0.32
1
2.84
9
1.21
9
0.64
17
10.23
69
6.64
85
11.55
43
10.98
51
10.94
60
13.59
55
11.65
52
4.30
33
8.94
44
4.63
37
0.11
58
0.04
51
0.00
1
0.00
1
0.05
16
0.46
76
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
PA-Nettwo views4.98
45
1.47
65
7.42
69
2.40
57
2.14
62
8.73
56
3.64
60
12.42
54
13.11
62
7.03
28
7.57
15
7.88
29
6.52
56
10.16
55
7.82
57
0.02
29
0.03
47
0.00
1
0.00
1
0.11
31
1.07
99
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
AANet_RVCtwo views5.01
46
1.74
77
6.38
56
1.96
41
1.29
43
2.26
2
1.69
43
10.07
22
18.53
83
7.88
33
18.15
74
8.49
35
2.70
15
10.59
59
7.04
51
0.96
105
0.15
83
0.02
56
0.00
1
0.13
35
0.12
41
PSMNet_ROBtwo views5.02
47
1.63
74
6.03
50
1.90
36
1.83
56
9.57
63
6.35
82
15.58
83
7.23
19
6.15
17
10.48
31
12.22
54
4.16
31
8.02
27
8.71
68
0.02
29
0.01
31
0.01
45
0.10
77
0.20
47
0.12
41
CBMVpermissivetwo views5.35
48
0.91
30
3.67
20
1.62
25
0.44
13
10.09
67
7.19
89
12.49
55
12.33
61
12.22
66
14.69
61
10.93
48
6.48
55
8.51
37
4.96
39
0.02
29
0.15
83
0.00
1
0.00
1
0.17
43
0.17
50
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
StereoDRNettwo views5.59
49
1.75
78
6.80
64
3.12
79
4.45
98
10.61
71
4.35
66
18.80
93
9.73
43
12.22
66
6.87
10
11.44
51
4.65
38
8.09
31
8.26
63
0.02
29
0.11
74
0.00
1
0.03
58
0.20
47
0.28
66
ETE_ROBtwo views5.80
50
1.77
79
6.33
54
1.44
17
0.78
27
6.43
45
6.90
86
12.53
56
8.08
32
12.93
74
14.89
62
21.13
94
5.87
52
9.83
51
6.57
49
0.04
45
0.01
31
0.00
1
0.02
49
0.08
23
0.33
68
DRN-Testtwo views5.87
51
0.98
36
5.89
48
2.69
67
3.65
90
12.37
77
3.35
54
20.07
104
10.20
48
11.93
65
12.31
49
11.06
50
5.31
48
7.89
25
9.05
70
0.04
45
0.05
55
0.04
70
0.04
67
0.18
45
0.25
61
NCCL2two views5.88
52
1.59
71
5.44
40
1.87
33
0.92
33
9.55
62
11.55
104
12.11
49
9.94
47
9.67
50
8.85
23
22.28
96
7.41
61
8.78
41
7.17
52
0.01
24
0.00
1
0.03
64
0.00
1
0.13
35
0.23
55
NaN_ROBtwo views6.00
53
1.24
53
6.29
53
1.34
12
1.68
52
9.60
64
10.31
100
15.09
76
15.79
74
12.62
69
8.95
24
11.67
53
5.83
51
11.78
67
6.41
48
0.05
49
0.13
80
0.08
77
0.20
84
0.22
55
0.79
92
DANettwo views6.02
54
1.23
52
8.45
74
3.86
94
3.94
92
7.64
53
1.34
39
9.51
18
7.00
16
13.39
76
15.53
65
15.99
73
7.02
60
12.14
68
12.37
86
0.19
69
0.12
79
0.02
56
0.03
58
0.13
35
0.56
83
XPNet_ROBtwo views6.03
55
1.22
51
5.61
44
2.56
64
0.90
31
6.32
42
7.07
87
12.92
59
8.30
35
14.76
81
15.13
64
19.84
89
6.66
59
10.36
56
8.58
66
0.02
29
0.04
51
0.00
1
0.03
58
0.11
31
0.24
58
Anonymous Stereotwo views6.16
56
3.15
99
23.75
108
2.97
77
2.48
72
4.39
27
13.30
106
9.21
15
9.86
45
9.56
49
8.76
22
6.79
19
1.99
9
13.50
78
13.04
89
0.01
24
0.05
55
0.00
1
0.06
71
0.22
55
0.19
52
GANettwo views6.22
57
1.07
42
4.07
26
2.27
53
0.89
29
9.19
58
9.52
95
12.02
48
8.13
33
10.72
57
29.09
98
13.86
61
7.52
63
11.00
63
4.39
34
0.36
83
0.00
1
0.02
56
0.02
49
0.12
33
0.08
27
DISCOtwo views6.28
58
0.57
12
5.78
46
3.43
86
1.17
39
11.22
73
3.39
55
12.14
51
16.16
76
6.52
24
11.22
36
16.96
76
6.32
54
19.51
99
10.74
80
0.00
1
0.00
1
0.00
1
0.00
1
0.35
69
0.11
40
RYNettwo views6.34
59
0.89
29
5.88
47
1.41
15
4.48
100
15.97
88
4.18
63
13.41
62
16.49
77
10.81
59
7.00
13
14.33
65
8.72
68
9.43
49
13.71
90
0.00
1
0.01
31
0.00
1
0.00
1
0.02
4
0.07
24
GANetREF_RVCpermissivetwo views6.56
60
2.89
94
7.58
71
3.41
84
0.40
11
12.96
80
9.58
96
15.09
76
17.25
82
10.33
54
10.62
33
12.27
55
8.16
65
12.21
69
4.53
36
0.41
87
0.00
1
0.00
1
0.02
49
3.12
111
0.39
71
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
LALA_ROBtwo views6.58
61
1.80
81
6.25
52
1.26
10
0.94
35
10.08
66
9.02
92
16.00
84
11.51
57
12.74
71
13.02
53
24.77
98
5.25
47
10.56
58
8.02
59
0.04
45
0.05
55
0.00
1
0.02
49
0.10
27
0.25
61
DeepPrunerFtwo views6.75
62
2.69
92
23.31
107
3.68
89
7.16
111
3.78
16
4.29
64
13.42
63
20.13
91
8.13
35
10.46
30
7.18
26
8.06
64
11.10
64
9.44
72
0.24
73
0.15
83
0.29
97
0.42
98
0.66
86
0.45
74
edge stereotwo views6.76
63
1.01
37
6.76
63
2.20
50
2.45
71
6.41
44
2.45
47
14.84
75
11.98
60
15.29
82
18.31
75
22.02
95
12.56
83
10.82
60
7.49
53
0.03
39
0.06
61
0.11
82
0.03
58
0.30
62
0.14
46
Abc-Nettwo views6.77
64
1.49
66
6.48
58
2.92
73
4.40
94
7.43
50
3.61
58
19.52
100
13.29
63
8.39
37
16.91
68
15.96
71
12.13
81
12.85
73
7.70
55
1.47
109
0.11
74
0.01
45
0.42
98
0.14
39
0.24
58
NCC-stereotwo views6.77
64
1.49
66
6.48
58
2.92
73
4.40
94
7.43
50
3.61
58
19.52
100
13.29
63
8.39
37
16.91
68
15.96
71
12.13
81
12.85
73
7.70
55
1.47
109
0.11
74
0.01
45
0.42
98
0.14
39
0.24
58
RPtwo views6.84
66
1.29
60
5.53
42
3.92
95
5.18
105
6.32
42
3.53
56
11.73
45
15.31
72
9.54
48
22.38
82
18.25
83
14.47
90
10.11
54
7.49
53
0.91
104
0.01
31
0.12
83
0.15
81
0.33
66
0.19
52
RGCtwo views6.88
67
2.23
88
6.13
51
4.05
96
4.73
104
8.94
57
2.78
49
15.19
80
11.74
59
11.13
61
19.34
77
17.86
80
10.42
76
13.02
75
8.03
61
0.73
98
0.01
31
0.24
96
0.41
97
0.31
64
0.38
70
Nwc_Nettwo views6.97
68
1.25
54
6.63
61
3.82
93
3.37
84
10.83
72
1.67
42
19.56
102
11.35
54
8.36
36
23.62
84
17.19
78
11.44
80
11.21
65
8.08
62
0.80
100
0.00
1
0.00
1
0.02
49
0.13
35
0.09
31
ADCReftwo views7.27
69
1.38
61
16.37
97
2.52
61
3.30
83
11.63
75
3.16
52
10.80
37
9.35
41
13.03
75
25.27
92
8.17
31
8.92
70
8.06
29
21.81
105
0.15
64
0.08
68
0.16
88
0.34
94
0.38
72
0.58
84
CSANtwo views7.62
70
1.60
72
6.56
60
1.83
31
0.66
18
12.40
78
10.52
102
14.45
72
21.32
93
14.19
78
15.98
67
17.84
79
13.02
87
12.32
70
8.38
64
0.09
55
0.07
66
0.03
64
0.04
67
0.33
66
0.67
90
stereogantwo views7.69
71
0.88
27
7.08
66
3.49
87
3.93
91
18.98
95
3.23
53
16.52
86
19.58
88
9.93
53
18.92
76
20.50
92
9.04
71
14.07
82
6.14
44
0.26
75
0.04
51
0.21
94
0.03
58
0.63
85
0.33
68
pmcnntwo views7.72
72
1.27
56
9.42
77
2.91
71
3.14
80
9.44
60
6.23
79
12.56
57
16.51
78
14.53
79
24.08
86
27.44
103
8.49
66
9.32
48
8.44
65
0.06
51
0.08
68
0.00
1
0.00
1
0.30
62
0.15
48
AF-Nettwo views7.78
73
1.44
64
6.68
62
3.37
83
4.50
101
8.61
55
2.69
48
17.07
89
20.17
92
9.52
47
24.02
85
20.31
91
14.59
91
11.58
66
9.84
77
0.61
96
0.00
1
0.12
83
0.00
1
0.38
72
0.12
41
PASMtwo views7.90
74
4.22
104
21.97
105
3.25
81
3.29
82
5.39
35
6.57
84
10.57
33
19.09
85
12.77
72
13.92
57
18.11
82
9.51
73
13.79
81
10.77
82
0.19
69
0.45
101
0.29
97
1.08
109
1.49
104
1.19
101
PWCDC_ROBbinarytwo views7.92
75
3.17
101
7.48
70
5.73
107
4.40
94
10.45
70
0.35
7
14.52
73
28.19
101
10.36
55
31.27
101
7.04
21
9.14
72
13.22
77
8.78
69
2.74
114
0.02
42
0.00
1
0.00
1
1.31
102
0.17
50
ADCP+two views8.09
76
1.79
80
14.50
95
1.54
20
4.28
93
16.57
90
5.20
74
12.80
58
11.20
53
12.83
73
17.07
70
11.02
49
10.80
78
17.59
95
23.18
108
0.03
39
0.05
55
0.01
45
0.18
82
0.39
76
0.81
93
SuperBtwo views8.10
77
3.15
99
24.67
109
2.65
65
1.23
40
9.88
65
4.29
64
10.18
28
30.07
104
11.53
63
12.18
46
6.12
16
6.65
58
10.50
57
14.47
92
0.14
62
0.11
74
0.35
101
0.25
88
13.06
118
0.48
77
PWC_ROBbinarytwo views8.24
78
3.13
97
12.74
89
2.43
59
4.43
97
7.51
52
1.22
34
16.63
87
19.24
86
16.08
84
28.29
95
13.99
64
10.16
75
13.63
80
14.06
91
0.42
90
0.00
1
0.05
73
0.00
1
0.59
82
0.27
65
MDST_ROBtwo views8.37
79
0.32
1
9.03
75
4.18
99
2.42
70
26.86
108
6.14
77
19.36
97
13.52
65
27.09
106
22.75
83
9.47
39
4.74
40
15.06
87
6.34
47
0.02
29
0.02
42
0.00
1
0.00
1
0.02
4
0.13
45
STTStereo_v2two views8.41
80
1.54
69
10.97
80
5.73
107
3.60
88
26.19
105
4.41
67
10.10
24
7.42
21
19.71
92
24.99
90
14.38
66
15.83
92
10.99
61
9.53
73
0.50
93
0.46
102
0.19
92
0.25
88
0.80
90
0.66
88
G-Nettwo views8.41
80
1.54
69
10.97
80
5.73
107
3.60
88
26.19
105
4.41
67
10.10
24
7.42
21
19.71
92
24.99
90
14.38
66
15.83
92
10.99
61
9.53
73
0.50
93
0.46
102
0.19
92
0.25
88
0.80
90
0.66
88
XQCtwo views8.43
82
3.58
102
16.40
98
2.92
73
2.17
63
13.22
82
3.60
57
14.64
74
25.86
98
11.87
64
12.04
44
15.06
69
10.67
77
15.24
88
19.41
97
0.39
85
0.08
68
0.05
73
0.07
73
0.84
92
0.45
74
FBW_ROBtwo views8.50
83
1.03
39
7.98
73
1.93
39
1.28
42
13.10
81
6.23
79
22.50
108
18.98
84
18.82
89
14.91
63
19.06
86
10.04
74
18.41
96
9.83
76
0.62
97
0.22
89
1.82
115
0.82
107
0.99
96
1.36
103
RTSCtwo views9.15
84
3.00
96
13.57
93
3.72
91
1.76
55
11.82
76
0.46
9
16.95
88
36.83
111
15.80
83
15.53
65
12.91
58
7.46
62
20.01
102
21.76
104
0.31
80
0.13
80
0.01
45
0.08
74
0.57
80
0.41
73
WCMA_ROBtwo views9.21
85
0.87
26
7.37
68
2.54
63
2.13
61
13.59
83
5.80
76
11.64
44
14.01
67
24.43
104
32.99
105
27.09
102
18.02
96
12.51
72
9.85
78
0.81
101
0.07
66
0.01
45
0.01
45
0.16
41
0.23
55
MSMD_ROBtwo views9.28
86
1.09
45
4.65
32
1.58
22
0.39
10
16.52
89
4.41
67
13.60
64
14.87
70
22.34
96
39.89
112
25.67
100
20.71
105
12.42
71
6.98
50
0.34
82
0.03
47
0.00
1
0.00
1
0.05
16
0.09
31
ADCPNettwo views9.54
87
2.39
90
31.46
112
2.09
44
1.60
48
16.71
92
6.39
83
12.11
49
11.45
55
13.53
77
21.45
80
19.41
87
10.94
79
14.38
84
21.54
103
0.27
78
1.16
109
0.39
104
1.49
113
0.58
81
1.45
104
SHDtwo views9.61
88
2.60
91
12.46
87
3.69
90
3.54
87
9.47
61
1.25
35
20.16
105
37.84
114
18.19
88
21.24
79
16.96
76
12.83
86
14.47
86
16.05
94
0.32
81
0.13
80
0.01
45
0.08
74
0.38
72
0.48
77
PDISCO_ROBtwo views9.62
89
1.99
85
11.51
82
9.88
115
9.61
116
21.48
99
3.83
61
19.33
96
28.49
102
11.27
62
14.17
58
19.92
90
5.02
45
16.35
92
9.18
71
5.28
116
0.41
97
0.14
87
0.09
76
2.05
108
2.36
112
MFN_U_SF_DS_RVCtwo views9.78
90
4.27
105
14.47
94
2.29
54
2.85
78
23.40
103
13.62
107
13.60
64
19.54
87
19.42
91
24.27
87
16.74
75
8.59
67
17.05
94
7.98
58
1.25
108
1.68
113
0.17
89
2.63
115
0.72
88
1.04
97
SGM_RVCbinarytwo views10.08
91
0.60
14
3.42
17
2.30
55
0.32
8
19.41
96
6.33
81
18.95
94
14.64
68
25.14
105
24.32
88
33.34
110
18.79
100
19.86
100
12.55
88
0.25
74
0.26
92
0.22
95
0.24
87
0.34
68
0.40
72
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
DPSNettwo views10.14
92
1.88
84
16.82
99
1.85
32
1.73
54
24.84
104
17.20
116
19.92
103
27.41
100
12.23
68
13.62
56
16.52
74
18.35
97
14.42
85
12.50
87
0.78
99
0.54
106
0.08
77
0.25
88
1.18
99
0.59
87
ADCLtwo views10.16
93
2.11
86
19.36
102
1.92
37
1.88
58
22.23
100
8.91
91
14.04
69
23.56
95
14.62
80
26.19
93
12.75
57
13.59
89
16.06
91
22.95
107
0.26
75
0.18
86
0.75
107
0.65
103
0.69
87
0.58
84
ADCMidtwo views10.24
94
3.13
97
20.70
103
2.21
51
2.39
69
11.23
74
6.19
78
14.17
70
11.19
52
23.20
102
22.25
81
17.89
81
19.54
102
18.51
97
26.21
111
0.45
91
0.42
99
1.10
111
1.29
110
1.56
106
1.18
100
SANettwo views10.64
95
1.86
83
10.91
79
1.76
29
0.71
21
14.62
86
9.23
94
19.18
95
37.14
112
19.22
90
27.96
94
25.86
101
19.11
101
13.02
75
10.63
79
0.08
54
0.06
61
0.03
64
0.02
49
0.62
84
0.81
93
FC-DCNNcopylefttwo views10.72
96
0.52
10
4.27
27
1.88
34
1.63
49
17.18
93
5.29
75
18.20
91
19.69
89
28.50
107
34.51
108
34.03
111
21.48
108
15.89
90
11.15
84
0.03
39
0.01
31
0.02
56
0.01
45
0.07
21
0.09
31
AnyNet_C32two views10.98
97
5.58
107
22.79
106
4.16
97
5.83
107
15.64
87
14.30
108
13.18
61
17.15
81
16.44
86
20.52
78
14.68
68
13.44
88
22.46
104
30.08
116
0.17
67
0.26
92
0.36
102
0.36
95
1.23
100
0.91
95
MeshStereopermissivetwo views11.52
98
1.52
68
4.55
31
1.89
35
1.46
44
19.87
98
5.11
73
20.66
106
15.91
75
32.67
112
34.51
108
39.34
116
21.15
106
18.74
98
12.10
85
0.11
58
0.06
61
0.01
45
0.00
1
0.45
79
0.22
54
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
MFN_U_SF_RVCtwo views12.94
99
3.66
103
25.81
110
3.61
88
2.26
66
22.77
101
4.55
71
27.10
113
20.06
90
23.90
103
28.99
97
30.53
107
16.98
94
19.92
101
20.26
99
1.24
107
1.07
108
0.98
110
1.33
111
1.80
107
2.04
108
ADCStwo views13.02
100
4.93
106
28.38
111
3.17
80
2.67
75
13.61
84
10.83
103
18.70
92
33.46
107
22.59
97
24.78
89
19.59
88
18.51
99
23.40
107
32.16
118
0.10
57
0.19
87
0.37
103
0.18
82
1.26
101
1.46
105
MFMNet_retwo views13.29
101
8.60
113
18.29
100
9.75
114
7.25
113
19.65
97
14.84
112
20.71
107
30.72
105
23.03
100
28.77
96
18.85
85
26.09
113
13.55
79
9.82
75
2.44
112
1.35
112
0.34
100
0.23
86
4.78
114
6.69
115
LSMtwo views14.01
102
5.95
108
33.49
113
6.78
112
43.61
120
10.22
68
9.98
99
15.16
79
22.93
94
23.07
101
32.34
104
18.52
84
12.67
84
15.45
89
11.10
83
0.16
66
0.51
105
0.09
80
0.32
92
1.08
97
16.85
120
SAMSARAtwo views14.63
103
2.74
93
12.38
86
12.65
117
6.74
110
36.50
114
72.93
122
19.36
97
23.77
96
16.20
85
13.04
54
29.21
104
12.78
85
16.98
93
15.21
93
0.11
58
0.26
92
0.03
64
0.14
80
0.76
89
0.77
91
SPS-STEREOcopylefttwo views15.04
104
6.23
109
13.21
92
11.34
116
11.65
118
23.30
102
7.15
88
24.16
110
15.65
73
31.78
111
29.19
99
31.62
108
21.32
107
24.62
108
19.50
98
7.59
118
4.19
119
3.22
116
1.48
112
6.99
117
6.54
114
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
PVDtwo views15.44
105
2.93
95
14.67
96
4.21
100
3.39
86
17.43
94
4.16
62
27.84
116
48.84
119
31.02
110
43.54
116
29.76
106
30.81
116
25.97
110
21.40
102
0.23
72
0.41
97
0.04
70
0.33
93
0.41
77
1.33
102
SGM+DAISYtwo views15.62
106
7.26
112
19.28
101
8.94
113
10.11
117
26.25
107
10.49
101
19.36
97
14.65
69
30.64
109
33.59
106
33.00
109
22.32
109
24.96
109
16.42
95
7.90
119
6.25
120
4.51
117
3.37
116
5.86
115
7.20
116
NVStereoNet_ROBtwo views16.04
107
6.75
110
12.90
90
6.37
111
7.42
114
12.89
79
9.74
97
22.78
109
25.12
97
30.32
108
46.19
118
34.37
112
25.38
111
21.48
103
21.38
101
5.94
117
3.10
117
6.07
118
10.09
120
4.01
112
8.54
118
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
AnyNet_C01two views16.12
108
10.81
117
59.36
118
4.42
101
2.49
73
30.06
110
15.15
114
17.51
90
16.51
78
17.88
87
37.69
111
24.04
97
17.54
95
29.60
113
33.29
119
0.28
79
0.38
95
0.43
105
0.42
98
2.57
110
1.98
107
MSC_U_SF_DS_RVCtwo views16.41
109
6.93
111
21.83
104
5.94
110
2.81
77
38.71
115
14.59
109
24.55
111
34.87
109
33.66
113
34.35
107
29.24
105
24.20
110
22.59
105
17.95
96
2.52
113
2.81
115
1.17
113
1.51
114
5.89
116
2.16
109
ELAS_RVCcopylefttwo views16.54
110
2.26
89
10.09
78
5.50
106
4.46
99
28.28
109
16.72
115
25.55
112
33.54
108
40.19
115
40.30
114
36.68
114
30.03
114
29.40
112
20.61
100
0.98
106
1.21
110
0.86
109
0.70
105
1.39
103
2.16
109
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
ELAScopylefttwo views16.72
111
2.14
87
9.23
76
4.92
103
4.53
102
32.66
113
15.11
113
27.40
114
28.68
103
40.27
116
44.90
117
38.33
115
30.50
115
26.44
111
21.94
106
0.88
102
1.23
111
0.67
106
0.89
108
1.49
104
2.18
111
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
LE_ROBtwo views16.73
112
1.28
59
11.61
83
3.72
91
1.65
50
16.67
91
9.17
93
14.39
71
55.91
120
63.81
120
40.86
115
35.94
113
37.73
119
14.24
83
26.87
112
0.05
49
0.10
72
0.13
86
0.22
85
0.12
33
0.15
48
SGM-ForestMtwo views16.99
113
1.08
44
5.74
45
2.12
46
0.75
23
31.63
112
12.21
105
27.80
115
32.25
106
37.88
114
39.99
113
52.96
119
35.20
118
33.60
115
24.47
109
0.26
75
0.39
96
0.31
99
0.39
96
0.26
61
0.53
82
DispFullNettwo views17.47
114
26.01
118
33.98
114
22.58
119
20.86
119
13.84
85
1.28
36
16.50
85
26.27
99
19.97
94
17.17
71
20.52
93
18.49
98
22.86
106
10.76
81
5.13
115
2.83
116
30.72
120
7.72
118
20.86
119
11.01
119
RTStwo views18.87
115
9.32
115
86.48
120
4.95
104
6.10
108
42.08
117
14.70
110
15.49
81
41.06
116
22.65
98
32.32
102
13.77
59
19.54
102
37.98
116
28.96
113
0.41
87
0.23
90
0.00
1
0.02
49
0.91
93
0.50
79
RTSAtwo views18.87
115
9.32
115
86.48
120
4.95
104
6.10
108
42.08
117
14.70
110
15.49
81
41.06
116
22.65
98
32.32
102
13.77
59
19.54
102
37.98
116
28.96
113
0.41
87
0.23
90
0.00
1
0.02
49
0.91
93
0.50
79
MANEtwo views19.47
117
1.27
56
5.07
36
4.69
102
5.55
106
30.49
111
9.94
98
34.01
118
37.27
113
44.13
117
51.57
121
52.51
118
40.41
120
33.58
114
24.81
110
0.89
103
0.86
107
1.11
112
9.72
119
0.38
72
1.06
98
BEATNet-Init1two views23.31
118
9.03
114
41.67
115
4.17
98
2.53
74
45.68
119
19.47
117
33.43
117
38.45
115
47.59
119
49.10
119
59.31
120
41.80
121
38.35
118
29.21
115
0.47
92
0.50
104
0.81
108
0.66
104
2.10
109
1.86
106
MADNet+two views27.07
119
33.84
119
90.97
122
20.14
118
7.47
115
48.43
120
47.10
119
35.43
119
36.46
110
20.11
95
30.05
100
25.29
99
35.08
117
45.50
120
50.28
120
2.13
111
2.00
114
1.19
114
0.76
106
4.71
113
4.43
113
PWCKtwo views30.53
120
44.32
120
47.25
117
29.76
120
7.23
112
40.78
116
27.10
118
44.73
120
44.32
118
47.31
118
36.37
110
47.16
117
26.05
112
41.26
119
31.87
117
21.83
120
4.03
118
29.50
119
4.67
117
27.17
120
7.80
117
DPSimNet_ROBtwo views53.45
121
64.73
121
44.39
116
53.97
121
45.39
121
53.66
121
54.83
120
55.15
121
57.87
121
64.16
121
50.83
120
63.40
121
53.34
122
46.45
121
65.81
121
63.13
121
26.54
121
57.94
121
51.11
121
45.52
121
50.69
121
MADNet++two views82.84
122
82.38
122
73.57
119
87.72
122
82.97
122
93.14
122
69.15
121
86.42
122
82.50
122
93.46
122
86.70
122
86.28
122
80.92
123
88.34
122
88.84
122
86.83
122
84.17
122
72.64
122
68.92
122
80.47
122
81.42
122
MEDIAN_ROBtwo views98.41
123
99.70
123
99.30
124
97.09
123
97.02
123
96.89
123
95.77
124
97.66
123
97.28
123
98.79
125
98.94
123
99.18
123
98.14
124
96.89
123
96.88
123
99.96
125
99.16
123
100.00
123
99.99
123
99.69
123
99.88
123
AVERAGE_ROBtwo views99.62
124
99.95
124
98.81
123
100.00
128
100.00
124
98.08
124
95.47
123
100.00
126
100.00
124
100.00
126
100.00
124
100.00
126
100.00
125
100.00
126
99.99
124
100.00
127
100.00
124
100.00
123
100.00
124
100.00
126
100.00
127
DGTPSM_ROBtwo views99.90
125
100.00
125
99.99
125
99.99
126
100.00
124
100.00
125
100.00
125
99.97
124
100.00
124
98.35
123
100.00
124
99.84
124
100.00
125
99.98
124
99.99
124
99.99
126
100.00
124
100.00
123
100.00
124
100.00
126
100.00
127
DPSMNet_ROBtwo views99.91
126
100.00
125
99.99
125
99.99
126
100.00
124
100.00
125
100.00
125
99.98
125
100.00
124
98.35
123
100.00
124
99.84
124
100.00
125
99.98
124
99.99
124
100.00
127
100.00
124
100.00
123
100.00
124
100.00
126
100.00
127
DPSMtwo views99.95
127
100.00
125
100.00
127
99.76
124
100.00
124
100.00
125
100.00
125
100.00
126
100.00
124
100.00
126
100.00
124
100.00
126
100.00
125
100.00
126
100.00
127
99.21
123
100.00
124
100.00
123
100.00
124
99.99
124
99.95
124
DPSM_ROBtwo views99.95
127
100.00
125
100.00
127
99.76
124
100.00
124
100.00
125
100.00
125
100.00
126
100.00
124
100.00
126
100.00
124
100.00
126
100.00
125
100.00
126
100.00
127
99.21
123
100.00
124
100.00
123
100.00
124
99.99
124
99.95
124
LSM0two views100.00
129
100.00
125
100.00
127
100.00
128
100.00
124
100.00
125
100.00
125
100.00
126
100.00
124
100.00
126
100.00
124
100.00
126
100.00
125
100.00
126
100.00
127
100.00
127
100.00
124
100.00
123
100.00
124
100.00
126
99.99
126
MSMDNettwo views1.26
6