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
R-Stereo Traintwo views0.44
1
0.09
4
0.94
1
0.02
8
0.00
1
1.24
2
0.09
23
3.65
19
0.11
1
1.87
22
0.42
13
0.06
1
0.01
1
0.25
1
0.05
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
19
R-Stereotwo views0.44
1
0.09
4
0.94
1
0.02
8
0.00
1
1.24
2
0.09
23
3.65
19
0.11
1
1.87
22
0.42
13
0.06
1
0.01
1
0.25
1
0.05
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
19
AdaStereotwo views0.65
3
0.09
4
1.12
4
0.02
8
0.00
1
1.45
18
0.23
40
4.38
43
1.44
22
1.17
15
0.37
11
1.56
13
0.18
12
0.56
3
0.37
6
0.00
1
0.00
1
0.00
1
0.00
1
0.01
5
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. ArXiv
CFNet_RVCtwo views0.77
4
0.46
27
1.34
8
0.01
3
0.27
59
1.32
7
0.02
9
2.40
1
0.68
4
0.63
4
0.18
3
3.23
43
0.21
16
3.97
16
0.54
14
0.00
1
0.00
1
0.01
67
0.00
1
0.03
15
0.02
19
DN-CSS_ROBtwo views0.77
4
0.90
68
2.01
21
0.85
76
0.00
1
1.47
20
0.01
6
2.92
4
0.93
5
0.12
1
0.53
17
0.44
4
0.16
10
4.33
21
0.35
4
0.00
1
0.00
1
0.00
1
0.00
1
0.30
76
0.01
8
HITNettwo views0.80
6
0.58
39
2.14
25
0.69
72
0.00
1
1.35
8
0.02
9
4.50
47
1.15
8
1.76
19
0.16
2
0.60
5
0.30
19
1.13
4
1.64
31
0.00
1
0.00
1
0.00
1
0.00
1
0.06
31
0.00
1
StereoDRNet-Refinedtwo views0.83
7
0.12
8
1.18
5
0.02
8
0.01
20
1.31
6
0.00
1
3.33
9
1.07
6
1.92
24
1.24
34
2.48
30
0.18
12
1.86
5
1.67
32
0.00
1
0.00
1
0.00
1
0.00
1
0.14
47
0.05
36
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
DeepPruner_ROBtwo views0.86
8
0.50
32
1.80
16
0.01
3
0.00
1
1.57
25
0.17
31
3.48
17
0.42
3
3.31
43
0.02
1
2.09
22
0.17
11
3.01
7
0.61
17
0.00
1
0.00
1
0.00
1
0.00
1
0.12
45
0.01
8
iResNettwo views1.00
9
0.45
26
3.08
42
0.86
79
0.01
20
1.97
35
0.01
6
3.67
21
1.49
28
1.77
20
1.09
26
0.35
3
0.13
8
3.83
14
1.27
27
0.00
1
0.00
1
0.00
1
0.00
1
0.07
33
0.01
8
MLCVtwo views1.01
10
0.53
34
2.17
26
0.03
14
0.00
1
1.60
28
0.03
12
2.90
3
1.30
11
2.60
34
1.85
43
1.00
7
0.40
24
4.88
28
0.74
21
0.00
1
0.00
1
0.00
1
0.00
1
0.07
33
0.01
8
ccs_robtwo views1.03
11
0.66
48
2.06
23
0.20
39
0.02
26
1.30
4
0.03
12
4.09
32
2.05
41
1.08
12
1.20
31
1.07
10
0.20
14
5.59
37
0.72
20
0.00
1
0.00
1
0.00
1
0.00
1
0.16
50
0.05
36
NLCA_NET_v2_RVCtwo views1.04
12
0.55
35
2.52
32
0.56
63
1.00
83
1.37
12
0.00
1
3.43
13
1.35
14
2.24
28
0.35
9
2.18
24
0.58
29
4.28
19
0.34
3
0.00
1
0.00
1
0.00
1
0.01
59
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.
CC-Net-ROBtwo views1.04
12
0.57
38
2.62
35
0.50
61
0.84
80
1.36
11
0.01
6
3.40
11
1.29
10
2.28
30
0.37
11
2.21
25
0.60
30
4.43
23
0.38
7
0.00
1
0.00
1
0.00
1
0.01
59
0.01
5
0.00
1
ccstwo views1.04
12
0.70
52
2.05
22
0.56
63
0.00
1
1.35
8
0.06
21
3.53
18
2.16
43
0.75
5
1.27
35
1.59
14
0.14
9
5.76
38
0.66
18
0.00
1
0.00
1
0.00
1
0.00
1
0.12
45
0.03
29
iResNet_ROBtwo views1.05
15
0.76
57
2.06
23
0.07
22
0.00
1
1.35
8
0.03
12
6.80
77
2.57
54
0.79
7
1.20
31
0.94
6
0.12
6
3.81
12
0.41
8
0.00
1
0.00
1
0.00
1
0.00
1
0.08
35
0.01
8
iResNetv2_ROBtwo views1.08
16
1.12
78
4.68
63
1.09
87
0.00
1
1.38
14
0.05
19
3.83
23
1.11
7
0.80
8
1.15
28
1.02
8
0.12
6
4.52
24
0.45
10
0.00
1
0.00
1
0.00
1
0.00
1
0.33
78
0.03
29
CFNettwo views1.08
16
0.62
44
2.26
27
0.26
45
0.03
29
1.40
15
0.04
17
4.45
46
2.09
42
0.48
2
0.69
20
1.31
11
0.24
17
6.63
53
0.83
22
0.00
1
0.00
1
0.00
1
0.00
1
0.17
53
0.04
33
TDLMtwo views1.08
16
0.49
30
1.43
10
0.27
46
0.00
1
1.56
24
2.12
64
3.10
5
1.75
35
1.36
17
0.93
24
1.06
9
0.49
26
6.21
44
0.70
19
0.00
1
0.00
1
0.00
1
0.00
1
0.16
50
0.02
19
NVstereo2Dtwo views1.09
19
0.17
9
4.40
61
0.00
1
0.00
1
1.59
27
0.18
34
4.29
40
1.40
19
0.85
9
0.36
10
2.30
27
0.62
31
5.26
32
0.35
4
0.00
1
0.00
1
0.00
1
0.00
1
0.01
5
0.01
8
PSMNet_ROBtwo views1.09
19
0.69
51
2.92
38
0.03
14
0.07
34
1.68
30
0.16
30
3.44
14
1.30
11
0.96
10
0.32
5
2.81
34
0.63
32
4.23
17
2.48
41
0.00
1
0.00
1
0.00
1
0.00
1
0.08
35
0.04
33
CVANet_RVCtwo views1.10
21
0.39
21
1.69
13
0.28
47
0.00
1
1.45
18
0.48
47
3.26
7
1.64
32
2.93
37
1.22
33
1.51
12
0.37
23
6.09
43
0.50
12
0.00
1
0.00
1
0.00
1
0.00
1
0.18
56
0.01
8
GANetREF_RVCpermissivetwo views1.10
21
0.95
69
2.98
39
0.06
19
0.00
1
2.23
36
0.18
34
3.95
29
2.53
52
0.56
3
0.32
5
1.89
19
0.66
36
5.03
29
0.51
13
0.00
1
0.00
1
0.00
1
0.00
1
0.05
28
0.04
33
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
ETE_ROBtwo views1.16
23
0.52
33
1.71
14
0.01
3
0.02
26
1.79
33
0.22
39
4.10
33
1.32
13
4.68
59
1.18
30
2.81
34
0.08
4
3.56
8
1.12
25
0.00
1
0.00
1
0.00
1
0.00
1
0.01
5
0.06
43
XPNet_ROBtwo views1.17
24
0.41
23
1.68
12
0.04
17
0.01
20
1.65
29
0.17
31
3.46
16
1.39
16
3.89
48
1.01
25
1.60
15
0.53
27
4.35
22
3.07
47
0.00
1
0.00
1
0.00
1
0.00
1
0.03
15
0.02
19
DLCB_ROBtwo views1.17
24
0.22
12
1.28
6
0.08
25
0.00
1
1.51
23
0.23
40
3.39
10
1.52
29
3.43
45
2.00
46
3.40
44
0.31
21
4.29
20
1.82
34
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
NOSS_ROBtwo views1.29
26
0.22
12
1.35
9
0.64
70
0.16
49
2.56
44
0.00
1
4.14
35
2.37
47
2.29
31
1.76
42
2.84
36
0.02
3
6.80
58
0.45
10
0.00
1
0.00
1
0.00
1
0.00
1
0.04
23
0.11
54
NCCL2two views1.30
27
0.63
45
2.31
28
0.02
8
0.03
29
1.57
25
3.41
73
3.91
28
1.39
16
3.14
41
0.34
7
3.90
48
0.30
19
3.75
10
1.19
26
0.00
1
0.00
1
0.02
75
0.00
1
0.03
15
0.05
36
PA-Nettwo views1.31
28
0.59
41
3.65
52
0.06
19
0.15
48
1.44
17
0.54
49
4.08
31
2.71
56
0.78
6
1.16
29
2.32
29
0.64
33
5.96
40
1.92
37
0.01
59
0.00
1
0.00
1
0.00
1
0.04
23
0.06
43
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
29
0.59
41
1.74
15
0.02
8
0.06
33
3.25
50
0.50
48
3.45
15
1.39
16
3.70
47
0.59
18
4.59
55
0.25
18
3.58
9
2.79
44
0.00
1
0.00
1
0.00
1
0.00
1
0.04
23
0.02
19
HSMtwo views1.36
30
0.40
22
1.07
3
0.01
3
0.30
61
1.73
32
0.03
12
5.89
69
1.22
9
1.83
21
1.36
36
4.78
57
1.30
41
6.61
52
0.57
15
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.00
1
HSM-Net_RVCpermissivetwo views1.40
31
0.07
1
1.58
11
0.00
1
0.01
20
1.88
34
0.05
19
7.24
78
1.71
34
3.63
46
3.63
56
3.21
42
0.36
22
4.25
18
0.44
9
0.01
59
0.01
53
0.00
1
0.00
1
0.01
5
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
32
0.75
54
4.58
62
0.05
18
0.10
41
2.49
43
1.10
57
4.62
49
1.47
26
3.38
44
0.62
19
2.29
26
0.65
35
4.52
24
2.89
45
0.00
1
0.03
64
0.00
1
0.01
59
0.05
28
0.07
47
RYNettwo views1.52
33
0.35
18
3.38
46
0.25
42
0.04
32
2.31
40
0.03
12
4.62
49
1.47
26
1.14
14
0.34
7
4.52
53
1.38
42
6.36
48
4.23
58
0.00
1
0.00
1
0.00
1
0.00
1
0.01
5
0.02
19
CBMVpermissivetwo views1.56
34
0.24
16
1.92
18
0.31
50
0.03
29
2.90
47
4.33
79
3.79
22
1.81
36
4.98
60
1.73
40
2.70
32
1.68
48
3.81
12
0.92
23
0.00
1
0.00
1
0.00
1
0.00
1
0.05
28
0.08
49
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
35
1.05
76
7.51
76
0.56
63
0.09
37
1.37
12
0.28
42
2.85
2
1.44
22
3.91
49
0.44
16
3.02
39
0.64
33
6.38
49
1.59
30
0.00
1
0.01
53
0.00
1
0.03
68
0.26
67
0.15
61
DRN-Testtwo views1.62
36
0.23
15
3.72
53
0.18
34
0.13
44
3.34
52
0.19
38
5.93
70
2.01
40
4.14
53
0.76
21
2.12
23
1.78
49
5.22
31
2.62
43
0.00
1
0.00
1
0.01
67
0.01
59
0.03
15
0.01
8
AANet_RVCtwo views1.66
37
0.88
66
3.52
50
0.06
19
0.00
1
1.22
1
0.00
1
3.25
6
2.51
51
2.21
27
4.49
58
3.65
45
0.20
14
7.10
64
3.42
52
0.50
92
0.04
68
0.00
1
0.00
1
0.08
35
0.05
36
NCC-stereotwo views1.67
38
0.56
37
3.10
43
0.08
25
0.22
53
2.24
37
0.13
28
5.05
61
1.42
21
2.27
29
2.72
47
4.29
49
3.53
66
5.76
38
2.01
39
0.04
73
0.00
1
0.00
1
0.00
1
0.03
15
0.02
19
CBMV_ROBtwo views1.70
39
0.11
7
1.98
19
0.46
59
0.00
1
2.85
46
0.91
53
4.27
38
2.21
46
4.02
51
5.00
62
2.97
37
2.32
54
5.40
33
1.35
28
0.00
1
0.00
1
0.00
1
0.00
1
0.03
15
0.06
43
DANettwo views1.74
40
0.22
12
3.06
41
0.75
74
0.44
68
2.46
42
0.35
45
3.40
11
1.41
20
4.22
54
1.56
37
5.03
59
0.56
28
6.29
46
5.00
64
0.01
59
0.00
1
0.00
1
0.01
59
0.04
23
0.08
49
RPtwo views1.76
41
0.55
35
2.59
34
0.22
41
0.60
78
1.70
31
0.98
55
4.31
42
1.99
39
3.12
40
5.04
63
4.41
51
2.96
58
4.72
27
1.90
35
0.01
59
0.00
1
0.00
1
0.01
59
0.14
47
0.01
8
Anonymous Stereotwo views1.78
42
1.37
84
7.79
77
0.42
56
0.07
34
1.49
22
3.20
71
3.31
8
1.84
37
2.75
35
1.14
27
1.84
17
0.09
5
6.89
61
3.32
50
0.00
1
0.00
1
0.00
1
0.00
1
0.09
41
0.01
8
SGM-Foresttwo views1.84
43
0.08
3
1.31
7
0.15
32
0.22
53
3.52
53
2.46
67
3.85
25
2.16
43
5.29
61
3.90
57
3.71
46
1.56
45
6.28
45
1.90
35
0.01
59
0.02
63
0.00
1
0.00
1
0.03
15
0.30
75
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
RGCtwo views1.86
44
0.67
49
3.45
49
0.07
22
0.26
57
3.29
51
0.11
26
4.78
55
1.46
25
3.26
42
4.51
59
5.14
60
2.68
57
5.46
34
1.94
38
0.00
1
0.00
1
0.00
1
0.00
1
0.08
35
0.09
52
GANettwo views1.90
45
0.43
24
1.83
17
0.39
54
0.00
1
2.24
37
3.37
72
3.90
27
1.44
22
2.17
26
9.66
72
2.57
31
1.09
38
6.88
60
1.81
33
0.00
1
0.00
1
0.02
75
0.00
1
0.10
42
0.03
29
DISCOtwo views1.93
46
0.18
10
2.44
30
0.67
71
0.53
74
5.25
65
0.00
1
5.43
64
1.56
30
2.09
25
0.91
23
6.48
67
0.89
37
8.55
74
3.60
55
0.00
1
0.00
1
0.00
1
0.00
1
0.06
31
0.01
8
G-Nettwo views2.00
47
0.70
52
4.12
58
0.03
14
0.21
52
2.28
39
0.11
26
7.33
80
1.67
33
2.34
32
3.43
54
5.31
63
3.62
68
5.50
35
3.38
51
0.01
59
0.00
1
0.00
1
0.00
1
0.03
15
0.02
19
MFMNet_retwo views2.02
48
0.98
70
4.27
59
2.74
91
0.88
81
1.47
20
0.15
29
4.21
36
3.18
60
6.63
65
6.93
67
1.87
18
3.63
69
2.13
6
0.99
24
0.00
1
0.00
1
0.00
1
0.00
1
0.23
64
0.05
36
NaN_ROBtwo views2.02
48
0.58
39
2.65
36
0.29
48
0.28
60
3.19
49
5.19
84
3.84
24
2.55
53
5.31
62
1.73
40
2.30
27
1.96
51
6.78
57
3.09
48
0.01
59
0.06
74
0.03
77
0.10
80
0.10
42
0.37
80
stereogantwo views2.13
50
0.31
17
3.75
54
0.01
3
0.09
37
7.44
78
0.60
50
4.27
38
3.29
61
3.91
49
4.62
60
6.41
66
1.84
50
5.17
30
0.58
16
0.00
1
0.01
53
0.01
67
0.00
1
0.24
65
0.05
36
PDISCO_ROBtwo views2.21
51
0.78
59
3.43
48
0.96
82
2.04
91
7.04
75
0.09
23
7.36
81
5.72
76
1.21
16
1.67
38
3.16
41
1.28
40
7.52
69
1.37
29
0.00
1
0.00
1
0.00
1
0.00
1
0.34
79
0.13
60
ADCReftwo views2.29
52
0.65
47
5.67
70
0.10
28
0.09
37
4.81
61
0.76
51
4.75
53
1.37
15
2.88
36
1.67
38
1.62
16
2.30
53
3.77
11
14.93
82
0.02
69
0.00
1
0.01
67
0.10
80
0.20
61
0.11
54
FBW_ROBtwo views2.36
53
0.36
19
3.16
44
0.37
53
0.14
45
4.42
60
0.18
34
6.64
74
3.05
59
2.98
38
2.85
48
4.86
58
1.11
39
11.00
85
4.53
59
0.16
83
0.05
72
0.37
89
0.09
77
0.19
60
0.78
89
DeepPrunerFtwo views2.50
54
0.87
65
14.90
87
0.17
33
0.31
62
1.41
16
0.29
43
5.62
65
9.97
82
1.02
11
0.43
15
2.01
20
1.65
47
6.85
59
3.99
56
0.00
1
0.01
53
0.07
79
0.10
80
0.24
65
0.12
59
XQCtwo views2.57
55
1.20
79
6.33
74
0.18
34
0.01
20
3.84
56
0.33
44
5.24
62
3.83
66
4.43
56
1.94
44
4.58
54
3.19
62
7.19
65
8.67
74
0.00
1
0.03
64
0.00
1
0.03
68
0.27
70
0.19
68
RTSCtwo views2.66
56
1.24
80
6.04
71
0.18
34
0.02
26
4.40
59
0.08
22
4.29
40
4.98
73
6.00
63
3.15
50
2.97
37
1.56
45
6.71
55
10.89
78
0.00
1
0.07
76
0.00
1
0.05
72
0.40
84
0.24
74
CSANtwo views2.70
57
0.75
54
3.41
47
0.18
34
0.12
42
4.18
58
4.05
76
4.11
34
4.36
70
6.03
64
6.24
66
4.60
56
4.74
73
7.29
66
3.43
53
0.02
69
0.03
64
0.01
67
0.02
67
0.18
56
0.22
71
ADCP+two views2.87
58
0.98
70
8.75
79
0.07
22
0.16
49
5.93
70
2.90
70
4.00
30
1.84
37
1.52
18
0.27
4
3.81
47
3.22
64
6.05
42
17.38
84
0.00
1
0.00
1
0.00
1
0.00
1
0.14
47
0.30
75
PWC_ROBbinarytwo views2.90
59
1.01
74
6.12
72
0.53
62
0.57
77
1.30
4
0.40
46
3.85
25
4.42
71
7.37
70
15.40
80
3.03
40
1.55
44
6.31
47
5.74
69
0.00
1
0.00
1
0.00
1
0.00
1
0.46
85
0.05
36
SHDtwo views3.01
60
1.02
75
5.52
68
0.45
58
0.39
65
4.12
57
0.17
31
6.68
75
12.04
84
7.20
69
3.36
52
4.40
50
2.58
56
4.69
26
7.17
72
0.01
59
0.04
68
0.00
1
0.03
68
0.22
63
0.17
66
PWCDC_ROBbinarytwo views3.06
61
1.38
85
3.81
55
0.14
31
0.00
1
3.14
48
0.02
9
4.40
44
12.35
85
1.08
12
23.69
89
2.01
20
2.01
52
3.87
15
2.43
40
0.24
89
0.00
1
0.00
1
0.00
1
0.61
91
0.08
49
SPS-STEREOcopylefttwo views3.08
62
0.49
30
2.49
31
0.21
40
0.24
55
3.67
55
1.24
58
5.80
68
2.43
50
10.68
79
8.00
68
9.48
76
3.59
67
8.11
72
5.08
65
0.01
59
0.01
53
0.00
1
0.00
1
0.04
23
0.11
54
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
NVStereoNet_ROBtwo views3.25
63
0.46
27
2.71
37
0.34
51
0.56
76
2.36
41
0.87
52
4.56
48
4.47
72
4.55
57
14.09
79
11.93
80
3.23
65
9.79
80
3.14
49
0.43
90
0.03
64
0.53
90
0.31
90
0.36
82
0.38
81
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
64
0.81
62
6.76
75
0.25
42
0.25
56
6.93
73
4.65
80
4.79
56
2.62
55
4.61
58
3.36
52
5.22
61
4.55
72
5.54
36
14.91
81
0.05
75
0.01
53
0.11
80
0.13
83
0.28
71
0.10
53
SAMSARAtwo views3.32
65
1.07
77
6.25
73
0.86
79
0.31
62
7.55
79
3.87
75
5.26
63
4.20
69
7.43
71
3.35
51
10.04
77
3.83
71
6.98
63
4.83
61
0.00
1
0.14
80
0.00
1
0.05
72
0.17
53
0.22
71
DPSNettwo views3.38
66
0.60
43
9.55
81
0.61
68
0.46
69
5.10
64
1.44
62
11.17
88
3.92
67
2.57
33
0.88
22
5.90
64
9.68
85
7.63
70
7.25
73
0.08
77
0.16
83
0.01
67
0.07
75
0.34
79
0.17
66
MDST_ROBtwo views3.50
67
0.07
1
3.94
56
2.66
90
1.46
87
12.94
87
0.97
54
7.32
79
2.42
49
14.82
88
8.08
69
2.74
33
1.51
43
8.44
73
2.48
41
0.00
1
0.00
1
0.00
1
0.00
1
0.01
5
0.11
54
pmcnntwo views3.61
68
0.68
50
5.38
67
0.61
68
1.46
87
2.75
45
1.26
59
5.04
60
1.57
31
8.77
75
11.67
73
18.69
89
2.38
55
6.03
41
5.61
68
0.00
1
0.07
76
0.00
1
0.00
1
0.21
62
0.07
47
ADCPNettwo views3.63
69
0.82
63
13.20
85
0.08
25
0.43
67
7.02
74
2.23
65
4.77
54
2.41
48
4.31
55
1.96
45
7.57
72
5.16
74
6.45
51
13.80
80
0.02
69
0.75
93
0.04
78
0.90
92
0.16
50
0.53
85
AnyNet_C32two views3.93
70
2.58
88
12.07
84
0.39
54
0.49
72
5.50
67
6.45
88
4.44
45
2.19
45
6.86
67
2.94
49
5.28
62
3.20
63
6.63
53
18.37
87
0.08
77
0.04
68
0.20
84
0.09
77
0.57
90
0.19
68
ADCMidtwo views3.95
71
1.34
82
10.91
83
0.18
34
0.38
64
5.02
63
1.27
60
4.87
57
2.94
58
11.15
80
3.60
55
6.76
69
5.21
76
6.40
50
16.96
83
0.12
80
0.08
78
0.55
91
0.46
91
0.53
89
0.19
68
SGM_RVCbinarytwo views4.07
72
0.38
20
1.98
19
0.84
75
0.26
57
6.20
71
2.37
66
6.45
72
3.57
63
11.56
82
9.27
71
15.55
86
6.78
78
10.07
82
4.60
60
0.22
84
0.25
87
0.20
84
0.23
87
0.26
67
0.34
78
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
SGM+DAISYtwo views4.39
73
1.34
82
5.18
65
1.01
84
0.88
81
5.61
68
2.86
69
4.88
58
3.53
62
13.29
84
13.95
78
12.66
82
7.55
79
7.98
71
5.39
67
0.23
88
0.18
85
0.15
82
0.24
88
0.29
72
0.64
86
WCMA_ROBtwo views4.54
74
0.48
29
3.19
45
0.49
60
0.63
79
5.38
66
2.57
68
4.70
51
3.77
65
14.17
86
19.02
84
15.47
85
9.03
82
6.73
56
4.85
62
0.01
59
0.05
72
0.01
67
0.01
59
0.08
35
0.16
62
SANettwo views4.63
75
0.89
67
5.34
66
0.30
49
0.09
37
7.43
77
3.75
74
6.53
73
16.16
88
6.79
66
13.29
75
12.44
81
7.74
80
7.38
67
4.11
57
0.00
1
0.01
53
0.00
1
0.00
1
0.10
42
0.31
77
MSMD_ROBtwo views4.94
76
0.43
24
2.58
33
0.12
29
0.01
20
8.43
82
1.09
56
4.25
37
4.11
68
13.11
83
26.36
91
13.95
84
13.88
89
7.45
68
2.92
46
0.10
79
0.01
53
0.00
1
0.00
1
0.01
5
0.02
19
ADCStwo views5.09
77
2.15
87
16.58
88
0.34
51
0.07
34
5.91
69
4.26
78
6.73
76
6.99
80
9.44
77
6.09
65
6.57
68
5.17
75
9.54
78
21.38
93
0.00
1
0.00
1
0.00
1
0.00
1
0.36
82
0.23
73
MeshStereopermissivetwo views5.78
78
0.75
54
3.04
40
0.25
42
0.14
45
7.88
81
2.09
63
9.64
84
5.18
74
20.48
90
19.06
85
22.80
91
9.54
84
9.64
79
4.88
63
0.00
1
0.00
1
0.00
1
0.00
1
0.29
72
0.03
29
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
79
0.99
73
5.56
69
0.72
73
0.51
73
6.72
72
0.18
34
9.90
86
23.69
93
11.45
81
19.26
86
8.18
73
9.46
83
9.24
76
9.49
75
0.02
69
0.16
83
0.00
1
0.08
76
0.18
56
0.40
83
Abc-Nettwo views5.94
80
1.31
81
9.31
80
0.58
66
0.14
45
12.38
86
6.50
89
15.85
91
6.41
77
7.14
68
17.35
81
11.01
79
14.09
90
9.80
81
6.06
70
0.00
1
0.01
53
0.00
1
0.00
1
0.50
87
0.39
82
FC-DCNNcopylefttwo views6.09
81
0.18
10
2.35
29
0.42
56
0.53
74
7.80
80
1.42
61
8.14
82
7.07
81
19.05
89
22.10
88
22.85
92
13.84
88
9.42
77
6.44
71
0.00
1
0.01
53
0.00
1
0.00
1
0.01
5
0.06
43
Nwc_Nettwo views6.24
82
0.98
70
8.69
78
0.13
30
0.17
51
10.25
85
8.78
92
16.70
92
5.49
75
4.08
52
13.66
77
10.29
78
10.74
87
16.71
89
17.65
85
0.05
75
0.00
1
0.00
1
0.01
59
0.32
77
0.16
62
AnyNet_C01two views6.59
83
4.31
91
36.24
93
0.96
82
0.46
69
9.96
83
5.74
85
6.02
71
3.63
64
7.59
72
8.56
70
8.83
74
3.14
61
13.74
87
20.83
92
0.12
80
0.08
78
0.25
87
0.09
77
1.01
92
0.34
78
LSMtwo views7.38
84
1.68
86
22.24
90
3.99
92
40.72
96
3.54
54
4.79
83
4.74
52
6.65
78
10.25
78
13.51
76
4.49
52
3.81
70
6.91
62
5.31
66
0.00
1
0.04
68
0.00
1
0.00
1
0.26
67
14.60
95
ELAS_RVCcopylefttwo views7.69
85
0.80
60
4.90
64
1.06
85
1.02
84
10.00
84
9.43
93
9.75
85
14.69
87
21.83
91
20.27
87
17.88
87
16.41
92
13.01
86
10.44
77
0.22
84
0.55
91
0.28
88
0.19
84
0.35
81
0.65
87
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
DispFullNettwo views7.91
86
23.96
96
13.67
86
14.71
95
7.88
94
4.83
62
0.04
17
5.01
59
2.89
57
8.86
76
4.96
61
6.11
65
10.72
86
10.18
83
3.51
54
2.44
93
0.33
90
15.15
95
4.65
95
12.33
95
5.95
93
ELAScopylefttwo views8.05
87
0.76
57
4.02
57
0.95
81
1.02
84
14.85
90
6.99
90
12.80
90
11.73
83
22.43
92
27.75
92
17.88
87
15.19
91
10.96
84
11.31
79
0.22
84
0.56
92
0.23
86
0.22
86
0.50
87
0.71
88
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
PWCKtwo views8.72
88
5.37
92
16.61
89
5.74
93
0.12
42
19.88
92
11.41
94
11.31
89
13.42
86
14.07
85
12.84
74
13.20
83
8.05
81
15.44
88
10.25
76
5.95
95
0.22
86
2.09
93
0.05
72
7.50
93
0.92
91
RTStwo views9.75
89
2.69
89
65.25
95
0.85
76
1.99
89
14.36
88
4.75
81
5.69
66
17.29
89
7.97
73
19.00
82
6.83
70
3.02
59
26.15
94
18.62
89
0.00
1
0.15
81
0.00
1
0.00
1
0.29
72
0.16
62
RTSAtwo views9.75
89
2.69
89
65.25
95
0.85
76
1.99
89
14.36
88
4.75
81
5.69
66
17.29
89
7.97
73
19.00
82
6.83
70
3.02
59
26.15
94
18.62
89
0.00
1
0.15
81
0.00
1
0.00
1
0.29
72
0.16
62
MADNet+two views10.71
91
14.57
94
70.07
97
0.58
66
0.47
71
19.89
93
4.20
77
11.05
87
6.87
79
3.03
39
5.99
64
9.01
75
6.35
77
31.95
96
28.31
95
0.22
84
0.26
89
0.01
67
0.04
71
0.46
85
0.87
90
SGM-ForestMtwo views12.56
92
0.80
60
4.31
60
1.07
86
0.41
66
21.56
94
8.57
91
17.42
93
19.67
91
29.93
93
28.49
94
45.83
96
27.22
95
26.05
93
18.40
88
0.14
82
0.25
87
0.18
83
0.28
89
0.18
56
0.42
84
LE_ROBtwo views12.79
93
0.64
46
10.26
82
2.13
88
1.29
86
7.25
76
5.81
86
9.09
83
49.27
96
57.55
96
24.58
90
23.68
93
31.13
97
9.00
75
23.57
94
0.04
73
0.06
74
0.13
81
0.20
85
0.08
35
0.11
54
MANEtwo views14.02
94
0.85
64
3.53
51
2.13
88
3.75
92
18.44
91
6.14
87
22.67
94
27.90
94
33.01
95
41.08
96
41.26
95
30.90
96
23.39
91
17.99
86
0.47
91
0.76
94
0.91
92
4.04
94
0.17
53
1.00
92
edge stereotwo views16.44
95
9.78
93
34.43
92
7.21
94
6.86
93
41.67
96
23.26
95
24.28
95
22.07
92
14.53
87
34.55
95
18.72
90
19.42
93
22.78
90
19.84
91
2.64
94
1.44
95
2.10
94
1.16
93
12.02
94
10.02
94
DPSimNet_ROBtwo views25.03
96
19.56
95
24.93
91
24.36
96
17.85
95
22.25
95
25.93
96
28.52
96
29.10
95
32.61
94
28.25
93
32.87
94
26.42
94
24.51
92
40.42
96
16.75
96
13.93
96
25.19
96
22.38
96
19.96
96
24.75
96
MADNet++two views47.78
97
34.15
97
37.11
94
42.94
97
41.23
97
64.94
97
30.63
97
60.79
97
53.26
97
72.08
97
61.65
97
57.68
97
54.96
98
58.03
97
59.24
97
43.01
97
36.17
97
36.07
97
22.50
97
47.50
97
41.67
97
MEDIAN_ROBtwo views96.83
98
99.41
98
98.66
99
94.75
98
94.23
98
93.08
98
90.54
98
95.61
98
94.75
98
97.65
100
97.73
98
98.30
98
96.57
99
94.51
98
93.74
98
99.78
103
98.24
98
99.99
100
99.89
98
99.48
100
99.67
100
DPSMtwo views99.17
99
100.00
102
100.00
102
96.28
99
98.84
99
100.00
100
100.00
100
100.00
101
100.00
99
100.00
101
100.00
99
100.00
101
100.00
100
100.00
102
100.00
102
93.66
98
100.00
99
100.00
101
100.00
99
96.31
98
98.27
98
DPSM_ROBtwo views99.17
99
100.00
102
100.00
102
96.28
99
98.84
99
100.00
100
100.00
100
100.00
101
100.00
99
100.00
101
100.00
99
100.00
101
100.00
100
100.00
102
100.00
102
93.66
98
100.00
99
100.00
101
100.00
99
96.31
98
98.27
98
AVERAGE_ROBtwo views99.23
101
99.81
99
97.56
98
100.00
104
100.00
101
96.24
99
91.02
99
100.00
101
100.00
99
100.00
101
100.00
99
100.00
101
100.00
100
99.98
101
99.97
99
100.00
104
100.00
99
100.00
101
100.00
99
100.00
104
100.00
103
DGTPSM_ROBtwo views99.45
102
99.94
100
99.98
100
96.58
101
100.00
101
100.00
100
100.00
100
99.81
99
100.00
99
95.86
98
100.00
99
99.32
99
100.00
100
99.79
99
99.99
100
98.34
100
100.00
99
99.86
98
100.00
99
99.52
101
99.99
102
DPSMNet_ROBtwo views99.46
103
99.94
100
99.98
100
96.59
102
100.00
101
100.00
100
100.00
100
99.81
99
100.00
99
95.87
99
100.00
99
99.32
99
100.00
100
99.79
99
99.99
100
98.46
101
100.00
99
99.86
98
100.00
99
99.52
101
100.00
103
LSM0two views99.94
104
100.00
102
100.00
102
99.84
103
100.00
101
100.00
100
100.00
100
100.00
101
100.00
99
100.00
101
100.00
99
100.00
101
100.00
100
100.00
102
100.00
102
99.22
102
100.00
99
100.00
101
100.00
99
99.99
103
99.83
101
MSMDNettwo views0.42
25