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 views2.12
1
0.98
11
3.94
8
0.20
1
0.11
1
4.73
21
0.13
2
9.27
9
4.40
4
6.46
13
5.03
4
0.45
1
0.72
1
3.79
3
1.94
4
0.00
1
0.01
22
0.00
1
0.00
1
0.05
7
0.24
13
PMTNettwo views2.13
2
0.37
1
2.87
1
0.43
2
0.25
6
5.01
23
0.48
9
6.49
1
8.65
24
3.32
1
3.45
1
0.74
2
0.91
4
8.32
18
1.20
1
0.00
1
0.00
1
0.00
1
0.00
1
0.03
4
0.07
1
R-Stereotwo views2.60
3
0.46
3
2.94
2
1.10
4
0.18
4
3.61
6
0.60
13
6.88
3
3.50
1
9.13
34
17.23
61
1.77
3
0.77
2
2.03
1
1.65
2
0.00
1
0.01
22
0.00
1
0.00
1
0.01
1
0.21
10
R-Stereo Traintwo views2.60
3
0.46
3
2.94
2
1.10
4
0.18
4
3.61
6
0.60
13
6.88
3
3.50
1
9.13
34
17.23
61
1.77
3
0.77
2
2.03
1
1.65
2
0.00
1
0.01
22
0.00
1
0.00
1
0.01
1
0.21
10
DN-CSS_ROBtwo views3.00
5
2.14
52
7.33
35
2.67
47
0.78
21
3.13
4
0.07
1
6.68
2
4.22
3
5.51
10
12.12
35
2.33
5
1.21
5
8.69
22
2.51
5
0.03
25
0.00
1
0.00
1
0.00
1
0.42
57
0.15
7
HITNettwo views3.11
6
1.38
21
5.35
19
2.23
28
0.12
2
5.59
28
0.58
11
9.95
10
5.72
7
6.62
15
7.91
10
3.90
6
3.24
14
4.39
4
4.58
23
0.01
13
0.00
1
0.00
1
0.00
1
0.08
12
0.57
32
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
ccstwo views3.26
7
0.58
5
3.71
5
2.06
25
0.53
15
2.06
1
0.49
10
13.61
52
14.34
57
4.14
5
5.97
6
5.46
11
2.40
10
5.25
5
3.12
10
0.09
42
0.08
54
0.12
69
0.10
59
0.22
36
0.94
63
AdaStereotwo views3.34
8
0.74
6
4.00
10
3.10
62
0.51
14
4.22
15
1.25
28
12.84
40
8.39
22
6.33
12
9.55
22
5.77
14
1.54
7
5.53
6
2.74
7
0.02
19
0.00
1
0.00
1
0.00
1
0.03
4
0.15
7
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 views3.66
9
1.78
36
6.24
28
1.79
18
2.84
62
2.19
2
0.63
16
8.56
6
4.87
5
8.81
28
9.30
20
9.12
31
4.94
32
7.14
8
4.31
19
0.04
29
0.10
58
0.06
63
0.03
40
0.24
40
0.23
12
NOSS_ROBtwo views3.67
10
0.86
9
3.81
6
2.53
43
1.11
31
6.10
31
0.72
20
11.08
24
12.36
45
5.45
9
8.91
19
5.65
12
2.09
8
8.30
17
3.16
11
0.20
54
0.07
53
0.00
1
0.00
1
0.11
20
0.80
56
BEATNet_4xtwo views3.69
11
2.12
51
8.11
44
1.82
21
0.16
3
5.30
25
1.04
25
10.87
21
6.14
8
7.09
20
8.42
17
4.99
9
4.42
25
5.89
7
6.23
32
0.04
29
0.05
44
0.00
1
0.00
1
0.32
51
0.84
57
CFNet_RVCtwo views3.70
12
1.59
33
3.87
7
1.68
13
2.42
57
3.20
5
0.66
18
8.92
7
7.76
14
5.09
6
9.95
23
10.77
40
5.43
40
7.41
10
4.71
25
0.00
1
0.00
1
0.03
55
0.01
32
0.30
46
0.11
2
MLCVtwo views3.76
13
1.52
28
6.93
33
1.66
12
0.31
7
4.31
17
0.32
6
7.92
5
7.68
13
9.55
39
12.12
35
5.17
10
2.74
12
10.00
40
4.49
22
0.01
13
0.00
1
0.00
1
0.00
1
0.16
25
0.27
15
DeepPruner_ROBtwo views3.82
14
1.87
39
5.65
24
1.31
7
1.64
41
3.62
8
0.81
23
14.47
58
5.03
6
8.03
25
10.78
27
6.95
18
2.26
9
8.92
26
4.32
20
0.07
39
0.03
37
0.00
1
0.01
32
0.37
55
0.29
16
STTStereotwo views3.90
15
1.43
22
8.20
46
3.03
58
2.28
55
3.64
9
0.64
17
10.11
11
7.09
10
10.20
41
7.08
7
8.73
30
3.57
16
7.44
11
4.00
14
0.00
1
0.02
30
0.01
40
0.01
32
0.04
6
0.40
18
ccs_robtwo views3.96
16
1.98
48
5.57
22
2.74
49
0.92
26
6.29
33
0.21
4
10.74
18
9.59
32
7.08
19
11.51
32
8.47
29
2.64
11
8.28
15
2.71
6
0.00
1
0.00
1
0.00
1
0.00
1
0.29
45
0.13
3
CFNettwo views4.03
17
1.97
47
6.30
29
2.70
48
1.64
41
5.41
26
0.22
5
12.14
34
10.28
33
5.24
8
11.15
30
6.43
17
3.70
18
8.97
29
4.01
15
0.00
1
0.00
1
0.00
1
0.00
1
0.30
46
0.13
3
iResNettwo views4.04
18
1.46
25
8.72
52
3.25
65
0.36
8
4.54
20
0.47
8
9.06
8
10.40
35
7.98
23
12.82
43
4.06
7
3.22
13
8.82
24
4.95
28
0.03
25
0.01
22
0.00
1
0.00
1
0.14
22
0.55
30
FADNet-RVC-Resampletwo views4.11
19
2.25
54
13.88
75
1.64
10
0.66
17
5.91
30
2.34
38
10.87
21
9.09
29
6.53
14
4.34
3
5.69
13
4.71
29
8.01
14
4.64
24
0.18
51
0.22
73
0.17
80
0.12
61
0.46
63
0.59
33
NLCA_NET_v2_RVCtwo views4.11
19
1.53
30
6.87
32
3.04
59
3.33
72
4.34
19
0.59
12
11.36
28
7.92
17
8.98
30
7.98
11
9.72
35
3.82
19
7.92
12
4.43
21
0.15
46
0.02
30
0.03
55
0.03
40
0.07
8
0.14
5
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 views4.12
21
1.54
31
6.86
31
2.98
56
3.04
66
4.18
14
0.67
19
11.10
25
8.07
20
9.05
32
8.35
16
9.47
32
3.99
21
7.94
13
4.73
26
0.14
45
0.02
30
0.03
55
0.03
40
0.07
8
0.25
14
FADNet_RVCtwo views4.24
22
2.32
60
14.23
78
1.18
6
0.79
22
5.64
29
0.60
13
11.50
29
7.19
11
3.68
2
8.08
14
4.14
8
4.06
22
10.29
43
7.47
37
0.36
69
0.44
82
0.18
81
0.47
87
0.93
83
1.23
71
HSMtwo views4.25
23
1.10
14
4.14
11
1.78
17
2.14
52
6.67
37
1.08
26
12.92
41
6.90
9
7.06
18
8.17
15
14.17
57
6.03
43
9.16
30
3.11
9
0.00
1
0.00
1
0.00
1
0.00
1
0.09
17
0.41
21
FADNet-RVCtwo views4.36
24
2.62
70
14.51
81
2.01
24
0.44
12
4.27
16
1.27
29
12.61
38
7.89
16
3.76
3
4.05
2
10.63
39
4.34
24
9.87
36
7.66
38
0.03
25
0.09
57
0.00
1
0.03
40
0.65
71
0.50
25
HSM-Net_RVCpermissivetwo views4.40
25
0.44
2
3.22
4
0.72
3
0.68
18
7.06
40
1.82
35
12.57
37
8.48
23
9.04
31
12.20
38
14.98
60
5.14
35
7.32
9
4.19
16
0.02
19
0.02
30
0.00
1
0.00
1
0.02
3
0.14
5
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
FADNettwo views4.55
26
2.35
61
13.60
74
1.87
22
0.80
23
4.74
22
0.75
21
14.47
58
12.41
46
4.07
4
5.45
5
9.64
34
5.03
33
8.86
25
4.82
27
0.20
54
0.04
40
0.07
64
0.05
50
1.27
90
0.40
18
TDLMtwo views4.58
27
1.79
38
5.03
16
2.45
39
1.32
38
4.04
12
7.15
76
11.60
31
11.46
41
6.91
17
12.69
42
5.94
15
5.03
33
9.74
33
3.80
13
0.67
85
0.00
1
0.01
40
0.05
50
0.23
38
1.78
87
CBMV_ROBtwo views4.66
28
0.85
8
3.94
8
1.64
10
0.75
20
8.38
45
1.90
36
11.19
26
10.36
34
9.20
38
15.21
51
7.28
21
6.99
49
9.48
31
5.24
30
0.04
29
0.05
44
0.00
1
0.00
1
0.07
8
0.64
40
iResNet_ROBtwo views4.67
29
1.77
35
5.93
26
2.47
40
0.95
27
3.78
10
0.36
7
15.65
66
17.36
69
8.75
27
11.23
31
7.05
20
4.60
27
8.94
27
4.23
17
0.01
13
0.00
1
0.00
1
0.00
1
0.16
25
0.15
7
CVANet_RVCtwo views4.68
30
1.93
44
5.57
22
2.61
45
1.86
48
4.12
13
4.94
60
12.11
32
9.29
30
8.05
26
11.08
29
10.28
37
6.48
46
9.52
32
3.19
12
0.45
73
0.01
22
0.00
1
0.12
61
0.23
38
1.75
86
iResNetv2_ROBtwo views4.72
31
2.26
56
8.87
55
3.20
64
1.25
36
4.32
18
1.73
34
14.22
56
10.62
37
10.53
45
12.13
37
7.77
24
5.54
41
8.29
16
2.76
8
0.00
1
0.00
1
0.00
1
0.00
1
0.42
57
0.55
30
NVstereo2Dtwo views4.82
32
1.29
18
7.98
43
3.68
74
3.48
74
8.02
42
3.10
43
11.25
27
15.50
61
5.14
7
7.71
8
8.21
27
4.68
28
10.19
42
4.30
18
0.02
19
0.19
71
0.13
72
0.36
82
0.07
8
1.12
69
StereoDRNet-Refinedtwo views4.84
33
0.96
10
5.13
17
2.22
27
0.46
13
9.10
48
0.16
3
10.72
17
9.40
31
13.41
64
12.50
40
10.02
36
4.58
26
8.60
20
8.71
45
0.01
13
0.00
1
0.01
40
0.05
50
0.24
40
0.53
28
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
DLCB_ROBtwo views4.87
34
1.45
24
4.75
15
2.52
42
1.13
34
6.25
32
3.01
41
10.44
16
8.34
21
11.05
49
13.43
45
14.04
54
3.90
20
9.74
33
6.80
34
0.00
1
0.00
1
0.00
1
0.00
1
0.08
12
0.37
17
RASNettwo views5.00
35
1.19
16
5.69
25
3.68
74
4.80
92
5.04
24
1.29
30
10.26
12
8.99
28
5.89
11
11.65
33
10.55
38
9.76
62
9.96
39
10.32
62
0.20
54
0.00
1
0.00
1
0.00
1
0.09
17
0.73
51
SGM-Foresttwo views5.40
36
0.84
7
4.19
12
1.43
8
0.68
18
10.12
58
6.41
68
12.13
33
11.86
43
11.50
51
14.43
49
11.62
44
5.20
36
9.93
38
6.16
31
0.16
48
0.04
40
0.00
1
0.00
1
0.08
12
1.24
72
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
PA-Nettwo views5.40
36
2.35
61
9.55
62
2.75
50
2.21
53
8.74
46
3.92
50
13.05
44
13.35
53
7.30
22
7.98
11
8.08
25
6.39
45
11.17
48
9.30
52
0.02
19
0.03
37
0.00
1
0.03
40
0.16
25
1.55
79
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
AANet_RVCtwo views5.41
38
2.61
69
8.17
45
2.25
30
1.30
37
2.27
3
1.64
32
10.82
19
18.54
75
8.01
24
18.09
66
8.39
28
3.31
15
11.37
51
8.98
49
1.15
92
0.38
78
0.02
49
0.00
1
0.17
29
0.70
48
PSMNet_ROBtwo views5.41
38
2.45
64
7.39
38
2.36
34
1.90
49
10.11
57
6.54
69
16.32
72
7.59
12
6.84
16
10.83
28
12.06
47
4.13
23
8.73
23
10.24
60
0.02
19
0.01
22
0.01
40
0.11
60
0.26
42
0.40
18
CBMVpermissivetwo views5.97
40
1.52
28
4.51
13
1.99
23
0.53
15
10.02
56
6.94
74
13.27
47
13.80
56
13.08
63
15.93
57
11.36
43
8.11
56
9.82
35
7.27
36
0.08
40
0.15
68
0.00
1
0.00
1
0.30
46
0.74
53
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
StereoDRNettwo views6.02
41
2.40
63
8.22
47
3.43
68
4.43
83
10.57
61
4.21
52
19.44
84
11.24
40
12.86
58
8.06
13
11.64
45
4.79
30
8.65
21
9.32
53
0.01
13
0.10
58
0.00
1
0.03
40
0.34
54
0.69
47
DRN-Testtwo views6.23
42
1.51
26
7.53
40
3.07
61
3.69
77
12.17
68
3.23
46
20.80
93
11.11
39
12.64
57
13.02
44
10.91
42
5.26
37
8.49
19
10.12
59
0.05
36
0.05
44
0.04
59
0.04
49
0.28
43
0.65
41
NCCL2two views6.27
43
2.46
65
7.00
34
2.36
34
1.11
31
9.48
51
11.14
88
13.01
43
10.40
35
9.94
40
9.49
21
22.11
88
7.42
52
9.90
37
8.56
42
0.01
13
0.00
1
0.02
49
0.03
40
0.20
32
0.73
51
ETE_ROBtwo views6.28
44
2.60
68
9.22
59
1.71
16
0.84
25
6.44
34
6.78
71
13.16
45
8.80
26
13.01
61
15.93
57
21.01
86
6.18
44
10.97
45
8.11
40
0.04
29
0.01
22
0.00
1
0.02
35
0.16
25
0.70
48
XPNet_ROBtwo views6.46
45
1.88
40
7.41
39
2.81
51
0.95
27
6.45
35
6.82
72
13.58
50
8.80
26
15.50
73
15.86
55
19.59
81
7.12
51
11.41
52
9.86
57
0.02
19
0.04
40
0.00
1
0.03
40
0.20
32
0.79
55
DANettwo views6.46
45
1.58
32
9.58
64
4.10
79
3.99
79
9.13
49
1.34
31
10.37
15
7.76
14
13.76
68
15.66
53
16.14
63
7.03
50
13.07
60
13.62
78
0.19
52
0.12
62
0.02
49
0.02
35
0.15
24
1.61
80
NaN_ROBtwo views6.51
47
2.25
54
8.77
53
1.69
14
1.79
46
9.71
53
10.94
86
16.20
70
16.13
63
13.04
62
8.86
18
11.77
46
5.94
42
12.88
59
8.01
39
0.05
36
0.13
65
0.07
64
0.20
69
0.32
51
1.50
76
DISCOtwo views6.58
48
0.98
11
6.44
30
3.75
76
1.18
35
11.19
63
3.29
48
12.92
41
16.72
65
7.27
21
11.74
34
17.34
68
6.60
47
19.59
91
11.59
70
0.00
1
0.00
1
0.00
1
0.00
1
0.45
62
0.50
25
RYNettwo views6.82
49
1.51
26
7.35
37
1.79
18
4.52
87
16.68
79
4.41
58
14.19
55
17.43
70
11.64
52
7.86
9
15.03
61
8.81
59
10.10
41
14.16
80
0.04
29
0.13
65
0.15
75
0.05
50
0.08
12
0.53
28
GANettwo views6.86
50
1.78
36
5.98
27
3.05
60
1.12
33
10.12
58
9.38
84
13.41
49
8.67
25
11.46
50
29.87
89
13.94
53
8.25
57
12.06
55
6.33
33
0.40
71
0.06
49
0.16
77
0.33
79
0.18
31
0.59
33
GANetREF_RVCpermissivetwo views6.97
51
3.59
83
9.14
58
4.00
77
0.40
9
13.17
70
11.27
90
15.95
68
17.51
71
10.62
46
10.66
26
12.15
48
7.97
55
13.18
61
5.04
29
0.45
73
0.00
1
0.01
40
0.15
64
3.47
102
0.60
35
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 views7.02
52
2.54
66
8.51
48
1.53
9
0.96
29
9.96
55
8.71
82
16.70
73
12.68
49
12.99
60
13.90
48
24.47
90
5.40
39
11.61
54
9.50
54
0.04
29
0.05
44
0.00
1
0.02
35
0.17
29
0.65
41
edge stereotwo views7.12
53
1.19
16
7.83
42
2.32
33
2.46
58
6.90
38
3.26
47
15.58
65
12.52
47
15.68
74
18.34
67
21.83
87
13.05
72
11.53
53
8.68
44
0.04
29
0.06
49
0.11
67
0.03
40
0.30
46
0.63
38
DeepPrunerFtwo views7.18
54
3.58
82
25.29
99
4.27
83
7.20
101
3.78
10
4.64
59
14.03
53
20.86
81
8.85
29
10.57
25
7.35
22
7.88
54
12.16
57
10.66
64
0.25
59
0.16
69
0.29
89
0.43
85
0.77
79
0.65
41
Anonymous Stereotwo views7.27
55
4.41
91
26.56
101
3.67
73
3.09
68
5.56
27
15.31
96
10.85
20
11.63
42
10.31
42
10.32
24
7.46
23
3.67
17
14.42
69
15.13
83
0.64
83
0.88
89
0.00
1
0.07
55
0.33
53
1.16
70
RGCtwo views7.40
56
2.77
72
7.71
41
4.43
86
4.87
93
8.88
47
2.96
40
16.05
69
12.76
50
11.96
55
19.57
69
18.32
73
11.57
69
13.89
67
9.58
56
0.83
88
0.02
30
0.24
88
0.41
84
0.40
56
0.77
54
RPtwo views7.41
57
1.90
42
7.33
35
4.26
82
5.25
94
6.64
36
3.46
49
12.44
36
16.23
64
10.36
43
23.00
74
18.72
74
16.59
84
11.10
46
8.47
41
1.10
91
0.02
30
0.12
69
0.16
65
0.43
59
0.65
41
Abc-Nettwo views7.43
58
2.28
57
8.58
49
3.28
66
4.48
84
8.34
43
4.27
55
20.45
91
13.66
54
9.17
36
17.52
63
16.49
64
14.31
79
13.71
64
8.95
47
1.73
97
0.12
62
0.01
40
0.50
89
0.21
34
0.60
35
NCC-stereotwo views7.43
58
2.28
57
8.58
49
3.28
66
4.48
84
8.34
43
4.27
55
20.45
91
13.66
54
9.17
36
17.52
63
16.49
64
14.31
79
13.71
64
8.95
47
1.73
97
0.12
62
0.01
40
0.50
89
0.21
34
0.60
35
Nwc_Nettwo views7.54
60
1.91
43
8.71
51
4.16
80
3.42
73
10.77
62
2.00
37
20.33
90
12.19
44
9.10
33
24.04
78
17.67
70
13.58
76
12.14
56
9.12
50
0.94
90
0.00
1
0.00
1
0.02
35
0.22
36
0.41
21
ADCReftwo views7.76
61
2.31
59
19.09
90
2.82
52
3.29
71
11.88
66
3.06
42
11.52
30
10.89
38
13.69
67
25.85
84
8.20
26
8.89
61
8.95
28
22.66
91
0.16
48
0.08
54
0.16
77
0.34
80
0.47
65
0.91
60
pmcnntwo views8.17
62
1.70
34
11.49
69
3.18
63
3.07
67
9.66
52
6.06
66
13.60
51
17.71
72
15.23
72
24.02
76
27.61
95
8.81
59
10.36
44
9.51
55
0.08
40
0.08
54
0.00
1
0.00
1
0.31
50
0.87
58
stereogantwo views8.18
63
1.32
20
9.05
57
3.59
70
4.00
80
20.16
88
3.13
44
17.25
78
20.16
79
10.68
47
19.19
68
20.56
85
10.70
64
14.74
73
7.00
35
0.34
67
0.04
40
0.21
86
0.08
56
0.70
75
0.65
41
SuperBtwo views8.29
64
3.96
87
24.92
97
2.96
55
1.33
39
9.83
54
4.15
51
10.92
23
29.79
94
11.86
53
12.51
41
6.07
16
6.97
48
11.13
47
14.73
81
0.15
46
0.11
60
0.35
91
0.26
74
12.64
109
1.11
68
AF-Nettwo views8.37
65
2.16
53
9.01
56
3.65
72
4.58
90
9.40
50
2.63
39
17.71
80
21.32
83
10.38
44
24.51
79
20.43
82
16.67
85
12.54
58
10.76
67
0.70
86
0.00
1
0.12
69
0.00
1
0.47
65
0.42
23
CSANtwo views8.41
66
2.86
75
9.35
61
2.43
38
0.82
24
13.59
72
11.68
91
15.80
67
21.98
84
14.71
70
16.46
59
18.00
72
14.42
81
13.68
63
10.10
58
0.10
43
0.13
65
0.15
75
0.19
67
0.43
59
1.28
73
PWCDC_ROBbinarytwo views8.50
67
4.40
90
9.57
63
6.16
99
4.54
89
10.46
60
1.03
24
15.30
62
28.16
91
10.88
48
30.88
92
6.96
19
11.31
66
14.57
71
10.40
63
3.10
105
0.02
30
0.00
1
0.00
1
1.55
94
0.63
38
G-Nettwo views8.60
68
1.93
44
12.59
71
6.04
96
3.66
75
25.93
96
4.26
53
10.35
13
8.05
18
19.71
83
24.73
80
14.15
55
15.49
82
11.30
49
10.69
65
0.50
80
0.46
83
0.19
83
0.25
72
0.80
80
0.91
60
STTStereo_v2two views8.60
68
1.93
44
12.59
71
6.04
96
3.66
75
25.93
96
4.26
53
10.35
13
8.05
18
19.71
83
24.73
80
14.15
55
15.49
82
11.30
49
10.69
65
0.50
80
0.46
83
0.19
83
0.25
72
0.80
80
0.91
60
ADCP+two views8.60
68
2.78
73
17.44
88
1.69
14
4.27
82
16.95
82
5.29
62
13.38
48
12.52
47
13.57
66
17.57
65
10.86
41
11.28
65
18.43
87
23.91
97
0.03
25
0.05
44
0.01
40
0.18
66
0.47
65
1.31
74
PWC_ROBbinarytwo views8.89
71
4.49
92
15.62
83
2.92
53
4.49
86
7.89
41
1.23
27
17.29
79
20.00
77
16.54
76
28.06
87
14.37
58
11.98
71
14.93
74
15.43
84
0.47
77
0.00
1
0.05
61
0.00
1
0.67
72
1.31
74
PASMtwo views9.11
72
5.68
96
25.06
98
4.33
84
4.18
81
7.01
39
8.48
81
12.37
35
20.95
82
13.49
65
15.37
52
18.80
75
11.52
68
14.70
72
12.70
75
0.93
89
1.19
95
0.32
90
1.21
99
1.66
95
2.22
93
MDST_ROBtwo views9.13
73
1.31
19
10.16
66
4.36
85
2.48
60
28.66
98
7.64
78
19.96
85
14.97
58
27.53
97
24.02
76
9.47
32
5.27
38
16.31
79
8.61
43
0.63
82
0.19
71
0.00
1
0.00
1
0.08
12
0.90
59
FBW_ROBtwo views9.17
74
2.08
50
10.31
67
2.51
41
1.66
44
13.22
71
6.04
65
23.09
100
20.11
78
19.55
81
15.80
54
18.97
78
11.96
70
19.34
88
11.38
69
0.66
84
0.30
77
2.15
106
0.90
95
1.15
88
2.16
91
XQCtwo views9.46
75
4.84
94
18.88
89
4.07
78
3.09
68
15.97
77
6.65
70
16.26
71
26.63
89
12.63
56
12.35
39
15.75
62
10.62
63
16.48
80
21.19
89
0.44
72
0.96
90
0.05
61
0.32
77
1.10
85
0.99
64
MSMD_ROBtwo views9.73
76
1.43
22
5.42
20
1.81
20
0.42
11
16.83
80
4.30
57
14.29
57
15.34
59
23.06
88
40.17
103
25.63
93
22.62
97
13.37
62
8.85
46
0.48
79
0.03
37
0.00
1
0.00
1
0.13
21
0.47
24
RTSCtwo views9.87
77
4.34
89
16.44
85
4.60
88
2.38
56
12.54
69
0.75
21
18.19
81
37.31
102
16.16
75
15.86
55
13.32
50
7.56
53
20.93
93
23.50
95
0.34
67
1.07
92
0.02
49
0.35
81
0.71
76
0.99
64
ADCPNettwo views10.18
78
3.40
80
33.24
103
2.38
36
1.74
45
17.75
84
7.65
79
13.20
46
13.03
52
14.09
69
21.72
71
19.36
79
11.40
67
15.43
76
23.00
93
0.28
64
1.15
94
0.39
94
1.48
102
0.68
73
2.21
92
WCMA_ROBtwo views10.20
79
1.99
49
9.33
60
3.01
57
2.46
58
15.78
76
7.67
80
12.75
39
15.41
60
25.08
95
33.42
96
27.06
94
19.86
89
13.87
66
12.47
74
1.28
94
0.38
78
0.18
81
0.20
69
0.28
43
1.54
77
PDISCO_ROBtwo views10.27
80
2.79
74
13.30
73
10.58
106
9.96
107
22.78
91
5.99
64
20.01
86
28.57
92
11.88
54
14.49
50
20.54
83
4.92
31
17.01
83
10.29
61
5.79
107
0.42
80
0.14
74
0.13
63
2.53
100
3.25
98
MFN_U_SF_DS_RVCtwo views10.36
81
5.37
95
16.61
87
2.95
54
3.02
65
24.75
93
15.44
98
14.68
60
19.73
76
19.77
85
23.97
75
16.98
66
8.40
58
17.58
85
9.23
51
1.59
96
1.75
99
0.16
77
2.63
106
0.75
77
1.85
88
SHDtwo views10.45
82
3.67
85
14.03
77
4.81
89
4.52
87
11.93
67
1.66
33
21.36
97
38.40
104
18.84
80
21.72
71
17.63
69
13.07
73
15.65
78
17.71
86
0.36
69
1.09
93
0.01
40
0.32
77
0.50
68
1.67
81
ADCLtwo views10.77
83
3.23
76
21.50
93
2.26
32
1.96
51
23.65
92
9.07
83
14.69
61
24.88
87
15.21
71
27.20
85
12.68
49
14.02
78
16.94
82
24.43
99
0.26
61
0.18
70
0.74
97
0.67
94
0.80
80
1.02
67
SGM_RVCbinarytwo views10.77
83
1.11
15
4.66
14
2.66
46
0.41
10
21.03
89
6.87
73
20.02
87
15.54
62
25.86
96
24.97
82
33.59
100
20.51
92
20.89
92
14.78
82
0.31
66
0.26
75
0.22
87
0.26
74
0.43
59
1.01
66
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
DPSNettwo views10.88
85
2.57
67
20.40
92
2.24
29
1.90
49
24.92
94
17.00
104
21.35
96
28.61
93
12.87
59
13.74
46
17.01
67
20.14
91
15.36
75
14.01
79
0.81
87
0.98
91
0.20
85
0.65
93
1.30
91
1.54
77
ADCMidtwo views10.97
86
4.49
92
23.66
95
2.53
43
2.52
61
11.46
65
6.31
67
15.31
63
12.85
51
23.57
92
22.99
73
17.92
71
21.04
93
19.57
90
28.05
102
0.47
77
0.43
81
1.10
100
1.32
101
1.73
97
2.10
89
FC-DCNNcopylefttwo views11.29
87
1.01
13
5.50
21
2.25
30
1.65
43
18.23
85
5.36
63
18.99
83
20.65
80
28.95
98
34.83
99
33.83
101
23.18
99
17.04
84
13.37
76
0.17
50
0.01
22
0.02
49
0.02
35
0.10
19
0.70
48
SANettwo views11.48
88
3.34
78
13.93
76
2.40
37
1.01
30
16.49
78
12.16
93
20.05
89
37.28
101
19.62
82
28.04
86
25.61
92
20.10
90
14.43
70
12.39
73
0.10
43
0.06
49
0.03
55
0.06
54
0.76
78
1.73
85
AnyNet_C32two views11.82
89
7.08
98
25.83
100
4.55
87
5.88
97
16.92
81
16.97
103
14.14
54
18.11
74
16.87
77
21.66
70
14.78
59
13.87
77
23.54
97
31.83
107
0.24
58
0.28
76
0.35
91
0.53
91
1.35
92
1.71
84
MeshStereopermissivetwo views11.94
90
1.88
40
5.24
18
2.11
26
1.44
40
20.09
87
4.95
61
21.74
99
16.75
66
33.10
103
35.07
100
39.28
107
22.72
98
19.45
89
13.59
77
0.19
52
0.06
49
0.02
49
0.00
1
0.46
63
0.65
41
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
ADCStwo views13.76
91
6.46
97
30.74
102
3.62
71
2.93
63
14.28
74
11.22
89
20.02
87
34.42
98
23.20
89
25.42
83
19.51
80
19.62
88
24.54
98
33.88
109
0.27
63
0.23
74
0.37
93
0.19
67
1.47
93
2.84
95
MFMNet_retwo views14.09
92
9.67
104
20.31
91
10.40
105
7.62
102
21.38
90
17.49
106
21.36
97
31.95
96
23.70
93
29.22
88
18.87
76
27.41
104
14.27
68
11.16
68
2.85
103
1.51
98
0.41
95
0.28
76
4.88
104
7.10
105
LSMtwo views14.85
93
7.41
100
36.29
105
7.76
103
43.77
111
11.24
64
12.14
92
16.73
74
23.72
85
23.73
94
32.92
93
18.89
77
13.31
75
16.60
81
12.37
72
0.26
61
0.86
88
0.11
67
0.43
85
1.24
89
17.26
111
SAMSARAtwo views15.67
94
3.89
86
14.48
80
13.77
108
7.92
104
38.69
105
73.90
113
21.22
95
24.66
86
17.04
78
13.86
47
30.01
97
13.07
73
18.07
86
17.48
85
0.23
57
1.46
96
0.07
64
0.40
83
0.99
84
2.15
90
SPS-STEREOcopylefttwo views15.83
95
7.38
99
14.92
82
11.63
107
11.77
109
25.30
95
7.12
75
25.20
102
17.06
68
32.24
102
30.34
90
31.63
99
22.59
96
25.67
99
21.97
90
7.71
109
4.50
109
3.35
107
1.56
105
6.97
108
7.72
106
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
PVDtwo views16.41
96
3.98
88
16.04
84
5.38
91
4.59
91
19.44
86
7.28
77
28.88
106
49.27
110
31.53
101
43.72
107
30.36
98
32.17
107
26.70
101
23.19
94
0.25
59
1.48
97
0.04
59
0.59
92
0.52
69
2.87
96
AnyNet_C01two views16.99
97
12.34
108
60.47
109
4.88
90
3.11
70
31.77
101
16.79
102
18.60
82
17.96
73
18.62
79
38.50
102
24.03
89
19.00
87
30.59
104
35.18
110
0.29
65
0.56
85
0.45
96
0.48
88
2.70
101
3.40
100
MSC_U_SF_DS_RVCtwo views17.08
98
7.98
101
24.37
96
6.82
101
2.93
63
39.68
106
16.11
99
25.46
103
34.93
99
34.13
104
34.20
97
29.20
96
24.94
101
23.01
95
19.39
88
2.79
102
2.87
105
1.18
102
1.53
104
6.78
107
3.31
99
NVStereoNet_ROBtwo views17.22
99
8.76
102
16.47
86
7.52
102
8.42
105
13.75
73
11.07
87
24.24
101
26.67
90
31.00
99
46.49
109
34.88
103
27.06
103
22.72
94
23.61
96
6.83
108
4.32
108
6.32
109
10.37
111
4.39
103
9.57
108
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
SGM+DAISYtwo views17.25
100
9.46
103
22.74
94
10.11
104
11.40
108
28.98
99
13.74
95
21.15
94
16.85
67
31.39
100
34.79
98
34.05
102
24.24
100
26.23
100
19.31
87
8.79
110
7.56
111
4.88
108
3.93
107
6.07
106
9.33
107
LE_ROBtwo views17.31
101
2.62
70
14.26
79
4.20
81
1.80
47
17.40
83
9.70
85
15.38
64
55.92
111
63.97
111
40.81
104
35.50
104
39.26
110
15.64
77
28.51
103
0.06
38
0.11
60
0.13
72
0.21
71
0.14
22
0.51
27
ELAS_RVCcopylefttwo views17.82
102
3.60
84
11.92
70
6.63
100
5.73
95
30.34
100
19.64
107
26.92
104
35.06
100
40.79
106
41.32
106
37.31
105
31.61
105
30.54
103
22.93
92
1.97
101
2.40
103
1.28
103
1.27
100
1.71
96
3.51
102
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
ELAScopylefttwo views17.99
103
3.49
81
11.19
68
6.14
98
5.79
96
34.89
104
17.28
105
28.75
105
30.47
95
40.80
107
45.79
108
39.04
106
32.05
106
27.63
102
24.30
98
1.88
100
2.41
104
1.09
99
1.48
102
1.83
98
3.48
101
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
DispFullNettwo views18.06
104
27.03
109
34.90
104
22.68
110
20.79
110
15.03
75
3.18
45
16.95
77
26.06
88
20.48
86
16.91
60
20.55
84
18.37
86
23.48
96
12.16
71
5.31
106
2.91
106
31.24
111
8.09
109
22.63
110
12.44
110
SGM-ForestMtwo views18.61
105
3.38
79
10.14
65
3.44
69
2.24
54
34.17
103
15.39
97
29.44
107
33.93
97
38.55
105
41.08
105
53.64
110
36.76
109
34.82
106
27.16
100
1.25
93
1.81
101
0.74
97
1.01
96
0.61
70
2.61
94
RTStwo views19.64
106
10.42
105
87.01
111
5.88
93
6.82
98
41.73
107
16.69
100
16.80
75
41.90
107
23.44
90
32.96
94
13.79
51
21.49
94
38.85
107
30.84
104
0.46
75
0.70
86
0.00
1
0.09
57
1.13
86
1.70
82
RTSAtwo views19.64
106
10.42
105
87.01
111
5.88
93
6.82
98
41.73
107
16.69
100
16.80
75
41.90
107
23.44
90
32.96
94
13.79
51
21.49
94
38.85
107
30.84
104
0.46
75
0.70
86
0.00
1
0.09
57
1.13
86
1.70
82
MANEtwo views20.98
108
3.23
76
8.81
54
5.96
95
6.96
100
33.07
102
13.20
94
35.43
109
38.86
105
44.72
108
52.41
112
53.20
109
41.84
111
34.74
105
27.45
101
1.87
99
2.27
102
1.54
105
10.18
110
0.69
74
3.11
97
BEATNet-Init1two views24.70
109
11.33
107
44.28
106
5.48
92
3.94
78
47.60
110
22.33
108
34.94
108
40.01
106
48.11
110
49.98
110
59.86
111
43.17
112
39.48
109
31.77
106
1.37
95
1.76
100
1.17
101
1.17
98
2.44
99
3.85
103
MADNet+two views27.96
110
35.32
110
91.41
113
21.04
109
8.86
106
49.25
111
47.41
110
36.98
110
37.84
103
21.00
87
30.83
91
25.41
91
34.95
108
46.51
111
51.54
111
3.04
104
3.33
107
1.48
104
1.05
97
5.85
105
6.05
104
PWCKtwo views31.62
111
45.59
111
49.67
108
30.49
111
7.80
103
42.91
109
29.73
109
45.45
111
44.44
109
47.41
109
36.92
101
47.86
108
26.79
102
41.90
110
33.36
108
22.77
111
4.56
110
29.89
110
5.13
108
29.43
111
10.36
109
DPSimNet_ROBtwo views54.45
112
65.73
112
47.00
107
54.49
112
46.37
112
55.34
112
56.47
111
56.18
112
59.04
112
64.57
112
51.76
111
64.01
112
54.49
113
47.54
112
67.21
112
63.21
112
27.86
112
58.04
112
51.28
112
46.61
112
51.87
112
MADNet++two views82.81
113
81.73
113
74.64
110
87.71
113
82.67
113
93.35
113
70.27
112
86.39
113
82.88
113
93.51
113
86.62
113
86.40
113
81.37
114
88.21
113
88.63
113
86.59
113
84.23
113
72.14
113
68.69
113
78.88
113
81.36
113
MEDIAN_ROBtwo views98.44
114
99.70
114
99.34
115
97.13
114
97.06
114
96.94
114
95.91
115
97.71
114
97.33
114
98.78
116
98.91
114
99.19
114
98.13
115
96.94
114
96.99
114
99.97
116
99.18
114
100.00
114
99.99
114
99.70
114
99.83
114
AVERAGE_ROBtwo views99.63
115
99.96
115
98.88
114
100.00
119
100.00
115
98.16
115
95.64
114
100.00
117
100.00
115
100.00
117
100.00
115
100.00
117
100.00
116
100.00
117
99.99
115
100.00
118
100.00
115
100.00
114
100.00
115
100.00
117
100.00
120
DPSMNet_ROBtwo views99.90
116
100.00
116
99.99
116
99.99
117
100.00
115
100.00
116
100.00
116
99.98
115
100.00
115
98.37
114
100.00
115
99.81
115
100.00
116
99.97
115
99.99
115
100.00
118
100.00
115
100.00
114
100.00
115
100.00
117
99.99
117
DGTPSM_ROBtwo views99.90
116
100.00
116
99.99
116
99.99
117
100.00
115
100.00
116
100.00
116
99.98
115
100.00
115
98.37
114
100.00
115
99.81
115
100.00
116
99.97
115
99.99
115
99.99
117
100.00
115
100.00
114
100.00
115
100.00
117
99.99
117
DPSMtwo views99.94
118
100.00
116
100.00
118
99.74
115
100.00
115
100.00
116
100.00
116
100.00
117
100.00
115
100.00
117
100.00
115
100.00
117
100.00
116
100.00
117
100.00
118
99.08
114
100.00
115
100.00
114
100.00
115
99.98
115
99.94
115
DPSM_ROBtwo views99.94
118
100.00
116
100.00
118
99.74
115
100.00
115
100.00
116
100.00
116
100.00
117
100.00
115
100.00
117
100.00
115
100.00
117
100.00
116
100.00
117
100.00
118
99.08
114
100.00
115
100.00
114
100.00
115
99.98
115
99.94
115
LSM0two views100.00
120
100.00
116
100.00
118
100.00
119
100.00
115
100.00
116
100.00
116
100.00
117
100.00
115
100.00
117
100.00
115
100.00
117
100.00
116
100.00
117
100.00
118
100.00
118
100.00
115
100.00
114
100.00
115
100.00
117
99.99
117
MSMDNettwo views1.28
6