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-Stereotwo views0.15
1
0.03
1
0.77
4
0.00
1
0.00
1
0.74
42
0.00
1
1.36
66
0.09
3
0.02
37
0.00
1
0.01
1
0.00
1
0.01
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.01
27
R-Stereo Traintwo views0.15
1
0.03
1
0.77
4
0.00
1
0.00
1
0.74
42
0.00
1
1.36
66
0.09
3
0.02
37
0.00
1
0.01
1
0.00
1
0.01
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.01
27
PMTNettwo views0.16
3
0.05
3
1.27
20
0.00
1
0.00
1
1.13
78
0.00
1
0.18
5
0.56
38
0.01
26
0.00
1
0.02
4
0.00
1
0.03
5
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.00
1
HITNettwo views0.19
4
0.55
70
1.36
23
0.00
1
0.00
1
0.25
5
0.00
1
1.09
53
0.23
18
0.00
1
0.00
1
0.14
16
0.02
14
0.02
4
0.14
33
0.00
1
0.00
1
0.00
1
0.00
1
0.05
43
0.00
1
Vladimir Tankovich, Christian Häne, Yinda Zhang, Adarsh Kowdle, Sean Fanello, Sofien Bouaziz: HITNet: Hierarchical Iterative Tile Refinement Network for Real-time Stereo Matching. CVPR 2021
AdaStereotwo views0.20
5
0.06
4
0.86
8
0.00
1
0.00
1
0.25
5
0.00
1
1.71
81
0.56
38
0.00
1
0.00
1
0.31
36
0.01
9
0.11
9
0.11
27
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.00
1
Xiao Song, Guorun Yang, Xinge Zhu, Hui Zhou, Zhe Wang, Jianping Shi: AdaStereo: A Simple and Efficient Approach for Adaptive Stereo Matching. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2021.
DPM-Stereotwo views0.22
6
0.10
11
0.77
4
0.00
1
0.00
1
1.09
74
0.00
1
1.68
79
0.57
43
0.01
26
0.00
1
0.01
1
0.07
47
0.01
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.01
27
MLCVtwo views0.23
7
0.46
51
1.15
17
0.00
1
0.00
1
0.45
18
0.00
1
0.57
24
0.15
9
0.09
61
0.00
1
0.27
30
0.03
25
1.09
18
0.25
41
0.00
1
0.00
1
0.00
1
0.00
1
0.05
43
0.00
1
DeepPruner_ROBtwo views0.23
7
0.37
41
1.26
18
0.00
1
0.00
1
1.11
77
0.03
65
0.20
7
0.07
1
0.00
1
0.00
1
0.39
46
0.03
25
1.12
19
0.02
10
0.00
1
0.00
1
0.00
1
0.00
1
0.04
39
0.00
1
StereoDRNet-Refinedtwo views0.24
9
0.08
9
0.68
2
0.00
1
0.00
1
0.83
48
0.00
1
0.66
27
0.61
60
0.00
1
0.00
1
0.50
53
0.02
14
1.28
21
0.08
23
0.00
1
0.00
1
0.00
1
0.00
1
0.10
67
0.04
64
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
STTStereotwo views0.24
9
0.28
24
1.68
34
0.00
1
0.00
1
0.84
49
0.03
65
0.16
4
0.08
2
0.02
37
0.00
1
0.20
23
0.03
25
1.54
22
0.01
5
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
MFMNet_retwo views0.24
9
0.74
88
2.31
57
0.53
105
0.00
1
0.17
2
0.01
46
0.02
1
0.59
56
0.01
26
0.03
59
0.12
12
0.00
1
0.07
6
0.01
5
0.00
1
0.00
1
0.00
1
0.00
1
0.19
88
0.01
27
BEATNet_4xtwo views0.24
9
0.51
62
1.71
38
0.00
1
0.00
1
0.31
7
0.00
1
0.99
47
0.31
21
0.00
1
0.00
1
0.29
32
0.03
25
0.22
11
0.35
47
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.02
41
iResNettwo views0.25
13
0.27
22
1.68
34
0.00
1
0.00
1
0.69
39
0.00
1
1.22
61
0.56
38
0.00
1
0.00
1
0.04
6
0.03
25
0.12
10
0.30
45
0.00
1
0.00
1
0.00
1
0.00
1
0.05
43
0.01
27
DN-CSS_ROBtwo views0.26
14
0.75
89
1.33
22
0.00
1
0.00
1
0.99
66
0.00
1
1.04
48
0.61
60
0.01
26
0.00
1
0.19
22
0.02
14
0.07
6
0.01
5
0.00
1
0.00
1
0.00
1
0.00
1
0.22
92
0.01
27
DLCB_ROBtwo views0.27
15
0.16
16
0.71
3
0.00
1
0.00
1
0.44
17
0.00
1
0.43
15
0.57
43
0.02
37
0.04
61
0.63
56
0.04
35
2.18
32
0.24
40
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
NCCL2two views0.29
16
0.50
57
1.52
28
0.00
1
0.00
1
0.81
46
0.00
1
0.45
16
0.22
15
0.00
1
0.00
1
0.02
4
0.03
25
2.14
30
0.06
19
0.00
1
0.00
1
0.02
93
0.00
1
0.01
11
0.03
53
ccstwo views0.29
16
0.08
9
0.79
7
0.00
1
0.00
1
0.88
53
0.22
76
1.71
81
0.63
77
0.00
1
0.01
39
0.30
35
0.23
67
0.86
15
0.03
12
0.01
85
0.00
1
0.03
98
0.01
82
0.06
49
0.02
41
iResNet_ROBtwo views0.30
18
0.66
80
1.41
25
0.00
1
0.00
1
0.70
40
0.00
1
0.79
34
0.57
43
0.04
51
0.01
39
0.08
9
0.01
9
1.55
23
0.03
12
0.00
1
0.00
1
0.00
1
0.00
1
0.06
49
0.00
1
DMCAtwo views0.30
18
0.38
42
1.10
16
0.00
1
0.07
88
0.87
52
0.03
65
0.04
2
0.14
7
0.29
72
0.17
69
0.04
6
0.37
73
2.26
35
0.13
30
0.00
1
0.04
92
0.01
85
0.01
82
0.01
11
0.04
64
DISCOtwo views0.31
20
0.11
12
1.73
39
0.16
95
0.16
94
0.31
7
0.00
1
0.09
3
0.17
11
0.00
1
0.01
39
1.51
84
0.00
1
1.21
20
0.62
60
0.00
1
0.00
1
0.00
1
0.00
1
0.03
34
0.01
27
CFNet_RVCtwo views0.31
20
0.34
32
0.96
13
0.00
1
0.00
1
1.08
72
0.00
1
0.19
6
0.15
9
0.01
26
0.00
1
0.48
51
0.00
1
2.95
49
0.05
17
0.00
1
0.00
1
0.01
85
0.00
1
0.01
11
0.01
27
ETE_ROBtwo views0.31
20
0.34
32
0.89
9
0.00
1
0.00
1
0.97
62
0.01
46
1.45
73
0.57
43
0.02
37
0.00
1
0.17
18
0.02
14
1.75
24
0.07
21
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
LALA_ROBtwo views0.31
20
0.38
42
0.91
12
0.00
1
0.01
55
1.00
67
0.01
46
1.60
75
0.61
60
0.15
67
0.00
1
0.09
11
0.01
9
1.04
17
0.41
50
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.00
1
NLCA_NET_v2_RVCtwo views0.35
24
0.49
56
2.00
46
0.00
1
0.02
63
0.96
60
0.00
1
0.55
23
0.18
14
0.03
45
0.00
1
0.35
42
0.10
52
2.37
36
0.01
5
0.00
1
0.00
1
0.00
1
0.00
1
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.
iResNetv2_ROBtwo views0.36
25
1.03
100
2.96
74
0.00
1
0.00
1
0.58
26
0.00
1
1.51
74
0.58
51
0.01
26
0.00
1
0.08
9
0.02
14
0.07
6
0.02
10
0.00
1
0.00
1
0.00
1
0.00
1
0.29
99
0.02
41
CC-Net-ROBtwo views0.37
26
0.50
57
2.09
54
0.00
1
0.02
63
0.98
63
0.00
1
0.46
17
0.17
11
0.08
58
0.00
1
0.33
37
0.11
54
2.54
41
0.01
5
0.00
1
0.00
1
0.00
1
0.01
82
0.01
11
0.00
1
XPNet_ROBtwo views0.37
26
0.34
32
1.01
14
0.01
54
0.00
1
0.96
60
0.00
1
1.11
55
0.46
28
0.00
1
0.00
1
0.17
18
0.02
14
2.73
44
0.54
57
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.01
27
PSMNet_ROBtwo views0.41
28
0.54
66
1.69
37
0.00
1
0.00
1
0.92
59
0.00
1
0.27
10
0.53
34
0.00
1
0.00
1
0.25
28
0.03
25
3.43
55
0.47
53
0.00
1
0.00
1
0.00
1
0.00
1
0.05
43
0.03
53
RPtwo views0.42
29
0.44
49
1.52
28
0.00
1
0.01
55
0.61
29
0.02
61
1.06
50
0.39
26
0.03
45
0.01
39
0.99
68
0.16
59
2.89
48
0.18
36
0.00
1
0.00
1
0.00
1
0.00
1
0.10
67
0.00
1
StereoDRNettwo views0.42
29
0.50
57
2.50
63
0.02
63
0.06
84
0.51
21
0.00
1
0.34
12
0.61
60
0.00
1
0.00
1
0.33
37
0.02
14
2.87
47
0.50
56
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.02
41
DRN-Testtwo views0.43
31
0.13
15
1.95
45
0.03
70
0.09
92
0.61
29
0.00
1
1.43
70
0.61
60
0.00
1
0.00
1
0.18
21
0.04
35
3.32
54
0.11
27
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.01
27
Nwc_Nettwo views0.43
31
0.42
48
2.64
67
0.00
1
0.00
1
0.67
37
0.00
1
0.82
38
0.28
19
0.00
1
0.01
39
0.90
64
0.10
52
2.62
42
0.05
17
0.00
1
0.00
1
0.00
1
0.00
1
0.03
34
0.00
1
NVstereo2Dtwo views0.45
33
0.11
12
1.91
44
0.00
1
0.00
1
0.59
28
0.00
1
1.39
68
0.61
60
0.03
45
0.00
1
0.33
37
0.00
1
3.94
64
0.07
21
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
ccs_robtwo views0.46
34
0.54
66
1.68
34
0.00
1
0.00
1
0.98
63
0.01
46
0.70
29
0.57
43
0.00
1
0.00
1
0.07
8
0.01
9
4.35
70
0.17
35
0.00
1
0.00
1
0.00
1
0.00
1
0.11
72
0.03
53
AF-Nettwo views0.47
35
0.53
64
2.05
51
0.00
1
0.00
1
0.41
13
0.01
46
1.18
60
0.40
27
0.13
63
0.05
62
1.30
78
0.45
74
2.43
40
0.30
45
0.00
1
0.00
1
0.00
1
0.00
1
0.09
63
0.00
1
stereogantwo views0.49
36
0.22
20
2.39
58
0.00
1
0.00
1
0.63
34
0.02
61
0.79
34
0.72
85
0.06
54
0.00
1
1.68
86
0.46
75
2.63
43
0.04
14
0.00
1
0.00
1
0.00
1
0.00
1
0.14
79
0.00
1
TDLMtwo views0.49
36
0.39
44
1.02
15
0.00
1
0.00
1
1.08
72
0.01
46
0.36
13
0.58
51
0.14
65
0.00
1
0.27
30
0.19
62
5.46
83
0.20
39
0.00
1
0.00
1
0.00
1
0.00
1
0.13
77
0.01
27
NCC-stereotwo views0.49
36
0.46
51
2.04
49
0.03
70
0.03
74
1.06
68
0.01
46
0.78
32
0.22
15
0.06
54
0.19
71
0.64
57
0.05
42
3.65
58
0.63
61
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.00
1
Abc-Nettwo views0.49
36
0.46
51
2.04
49
0.03
70
0.03
74
1.06
68
0.01
46
0.78
32
0.22
15
0.06
54
0.19
71
0.64
57
0.05
42
3.65
58
0.63
61
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.00
1
RGCtwo views0.50
40
0.53
64
2.42
59
0.00
1
0.01
55
0.65
35
0.00
1
1.07
51
0.17
11
0.01
26
0.62
81
1.00
69
0.19
62
3.03
50
0.28
44
0.00
1
0.00
1
0.00
1
0.00
1
0.05
43
0.02
41
SHDtwo views0.50
40
0.51
62
3.61
87
0.05
79
0.00
1
0.01
1
0.01
46
0.72
30
0.61
60
1.57
90
0.00
1
0.34
40
0.09
49
1.79
26
0.49
55
0.00
1
0.00
1
0.00
1
0.01
82
0.09
63
0.03
53
PASMtwo views0.51
42
0.32
30
3.07
77
0.00
1
0.00
1
0.31
7
0.00
1
0.46
17
0.12
5
0.03
45
0.00
1
0.44
48
0.02
14
5.22
79
0.04
14
0.00
1
0.00
1
0.00
1
0.00
1
0.12
74
0.03
53
FADNet-RVCtwo views0.51
42
0.80
93
2.47
62
0.01
54
0.00
1
0.48
20
0.00
1
1.32
64
0.14
7
0.00
1
0.01
39
0.12
12
0.06
46
4.07
66
0.15
34
0.00
1
0.04
92
0.00
1
0.00
1
0.38
102
0.08
83
FADNet-RVC-Resampletwo views0.51
42
0.82
94
3.08
78
0.04
75
0.04
80
0.42
15
0.09
72
1.16
58
0.34
22
0.00
1
0.03
59
0.24
27
0.05
42
3.16
51
0.13
30
0.07
100
0.12
102
0.05
101
0.02
90
0.08
60
0.20
101
HSM-Net_RVCpermissivetwo views0.52
45
0.06
4
1.45
26
0.00
1
0.00
1
0.61
29
0.00
1
3.83
106
0.52
32
0.14
65
0.00
1
0.22
26
0.03
25
3.50
57
0.04
14
0.01
85
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
CVANet_RVCtwo views0.53
46
0.32
30
1.26
18
0.00
1
0.00
1
0.98
63
0.00
1
0.72
30
0.55
37
0.35
74
0.12
68
0.36
44
0.14
56
5.53
85
0.08
23
0.00
1
0.00
1
0.00
1
0.00
1
0.16
84
0.00
1
RYNettwo views0.53
46
0.19
18
1.79
41
0.00
1
0.00
1
0.55
24
0.00
1
1.12
56
0.61
60
0.02
37
0.00
1
0.83
62
0.04
35
5.05
77
0.45
52
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
41
GANetREF_RVCpermissivetwo views0.54
48
0.70
83
2.06
52
0.00
1
0.00
1
1.14
79
0.00
1
1.61
76
0.58
51
0.00
1
0.01
39
0.29
32
0.04
35
4.21
68
0.06
19
0.00
1
0.00
1
0.00
1
0.00
1
0.03
34
0.02
41
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
FADNettwo views0.55
49
0.75
89
2.42
59
0.00
1
0.01
55
0.82
47
0.00
1
0.88
40
0.54
36
0.01
26
0.01
39
0.16
17
0.16
59
4.49
71
0.09
25
0.18
104
0.00
1
0.00
1
0.00
1
0.43
105
0.00
1
GANettwo views0.55
49
0.35
37
1.40
24
0.00
1
0.00
1
0.77
45
0.08
71
0.36
13
0.57
43
0.01
26
0.00
1
0.68
59
0.11
54
5.52
84
1.07
66
0.00
1
0.00
1
0.02
93
0.00
1
0.09
63
0.03
53
PDISCO_ROBtwo views0.56
51
0.57
72
2.07
53
0.17
96
0.00
1
1.56
87
0.01
46
2.27
94
0.51
31
0.08
58
0.01
39
0.74
60
0.28
70
2.37
36
0.13
30
0.00
1
0.00
1
0.00
1
0.00
1
0.33
101
0.00
1
CFNettwo views0.56
51
0.50
57
1.58
30
0.00
1
0.00
1
1.06
68
0.01
46
1.08
52
0.57
43
0.00
1
0.01
39
0.12
12
0.01
9
5.87
93
0.19
38
0.00
1
0.00
1
0.00
1
0.00
1
0.13
77
0.02
41
CBMVpermissivetwo views0.57
53
0.22
20
1.61
32
0.00
1
0.00
1
0.84
49
0.78
95
2.58
98
0.61
60
0.38
76
0.00
1
1.25
75
0.63
79
2.07
29
0.38
48
0.00
1
0.00
1
0.00
1
0.00
1
0.04
39
0.08
83
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
PA-Nettwo views0.57
53
0.50
57
2.73
71
0.03
70
0.06
84
0.89
55
0.02
61
0.66
27
0.63
77
0.00
1
0.00
1
0.43
47
0.07
47
4.82
73
0.55
59
0.01
85
0.00
1
0.00
1
0.00
1
0.02
22
0.04
64
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
DANettwo views0.58
55
0.19
18
1.87
42
0.26
101
0.06
84
0.51
21
0.11
73
1.41
69
0.52
32
0.22
69
0.06
65
1.06
71
0.02
14
4.25
69
0.92
64
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.06
77
RASNettwo views0.58
55
0.27
22
1.30
21
0.00
1
0.00
1
0.61
29
0.01
46
0.47
19
0.60
58
0.08
58
0.06
65
1.52
85
0.04
35
6.18
96
0.48
54
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.00
1
edge stereotwo views0.58
55
0.29
28
2.86
72
0.01
54
0.03
74
0.33
10
0.03
65
0.97
45
0.50
29
0.21
68
0.33
76
1.31
79
0.16
59
4.20
67
0.18
36
0.00
1
0.00
1
0.00
1
0.00
1
0.11
72
0.01
27
Anonymous Stereotwo views0.59
58
0.78
92
2.67
68
0.00
1
0.00
1
0.55
24
0.49
87
1.05
49
0.28
19
0.00
1
0.01
39
0.12
12
0.02
14
5.54
86
0.27
43
0.00
1
0.00
1
0.00
1
0.00
1
0.07
54
0.00
1
FADNet_RVCtwo views0.60
59
0.83
95
2.50
63
0.00
1
0.00
1
1.09
74
0.00
1
1.29
62
0.50
29
0.02
37
0.05
62
0.26
29
0.03
25
3.17
53
0.25
41
0.21
106
0.26
106
0.08
105
0.28
108
0.53
107
0.55
107
NOSS_ROBtwo views0.61
60
0.17
17
0.63
1
0.19
98
0.00
1
0.67
37
0.00
1
2.38
95
0.57
43
0.13
63
0.00
1
0.94
65
0.00
1
6.27
99
0.10
26
0.00
1
0.00
1
0.00
1
0.00
1
0.04
39
0.10
90
PWC_ROBbinarytwo views0.62
61
0.54
66
3.22
80
0.00
1
0.00
1
0.22
4
0.00
1
2.18
92
0.61
60
0.00
1
0.20
73
0.29
32
0.03
25
3.46
56
1.22
71
0.00
1
0.00
1
0.00
1
0.00
1
0.39
103
0.01
27
HSMtwo views0.63
62
0.36
40
0.89
9
0.00
1
0.00
1
0.52
23
0.01
46
0.98
46
0.56
38
0.00
1
0.01
39
2.27
93
0.65
80
6.18
96
0.11
27
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.00
1
XQCtwo views0.68
63
0.72
87
3.82
90
0.06
84
0.00
1
0.39
12
0.01
46
1.92
86
0.61
60
0.03
45
0.01
39
0.46
50
0.21
64
2.84
46
2.26
82
0.00
1
0.00
1
0.00
1
0.01
82
0.15
81
0.05
69
RTSCtwo views0.70
64
0.71
85
3.19
79
0.01
54
0.00
1
0.20
3
0.00
1
0.81
36
0.60
58
0.29
72
0.36
78
0.36
44
0.04
35
2.04
28
4.87
93
0.00
1
0.03
88
0.00
1
0.02
90
0.28
97
0.11
91
PVDtwo views0.73
65
0.57
72
3.38
84
0.04
75
0.19
99
0.74
42
0.00
1
0.29
11
0.79
88
2.37
94
0.00
1
0.53
54
0.09
49
2.19
33
3.20
87
0.00
1
0.05
98
0.00
1
0.02
90
0.08
60
0.15
98
AANet_RVCtwo views0.75
66
0.65
78
2.52
65
0.01
54
0.00
1
0.66
36
0.00
1
0.61
25
0.13
6
0.00
1
1.29
90
0.96
66
0.04
35
5.98
94
1.82
78
0.13
101
0.02
85
0.00
1
0.00
1
0.07
54
0.04
64
NaN_ROBtwo views0.80
67
0.47
54
1.63
33
0.17
96
0.19
99
0.58
26
0.38
83
2.11
89
0.80
89
0.53
77
0.33
76
0.48
51
1.13
91
4.98
76
1.94
79
0.00
1
0.02
85
0.01
85
0.05
97
0.06
49
0.14
97
FBW_ROBtwo views0.83
68
0.31
29
1.48
27
0.10
90
0.01
55
0.62
33
0.01
46
1.91
85
0.59
56
0.01
26
0.01
39
1.10
73
0.05
42
7.84
104
2.07
81
0.05
98
0.01
82
0.03
98
0.01
82
0.01
11
0.47
105
CBMV_ROBtwo views0.84
69
0.06
4
1.73
39
0.00
1
0.00
1
1.37
83
0.47
84
2.57
97
0.58
51
2.13
92
0.76
84
1.06
71
1.38
93
4.04
65
0.54
57
0.00
1
0.00
1
0.00
1
0.00
1
0.03
34
0.05
69
SGM-Foresttwo views0.84
69
0.07
8
0.90
11
0.02
63
0.07
88
1.10
76
0.68
91
2.12
91
0.61
60
1.14
85
0.78
85
1.36
80
0.98
87
5.55
87
1.31
74
0.00
1
0.00
1
0.00
1
0.00
1
0.02
22
0.06
77
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
SPS-STEREOcopylefttwo views0.88
71
0.40
46
2.01
47
0.00
1
0.01
55
0.42
15
0.17
74
0.50
21
0.61
60
2.19
93
0.79
86
2.04
91
0.27
69
5.10
78
2.89
86
0.00
1
0.00
1
0.00
1
0.00
1
0.03
34
0.11
91
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
ADCReftwo views0.92
72
0.48
55
3.35
82
0.02
63
0.02
63
0.90
56
0.32
81
0.61
25
0.65
81
0.72
78
0.17
69
0.17
18
0.82
85
2.15
31
7.95
97
0.00
1
0.00
1
0.00
1
0.00
1
0.07
54
0.03
53
PWCDC_ROBbinarytwo views0.94
73
0.91
98
2.52
65
0.00
1
0.00
1
1.15
80
0.00
1
2.11
89
1.18
95
0.00
1
8.99
102
0.35
42
0.09
49
0.64
13
0.40
49
0.00
1
0.00
1
0.00
1
0.00
1
0.50
106
0.03
53
SAMSARAtwo views1.01
74
0.70
83
3.29
81
0.20
99
0.06
84
1.65
88
0.05
70
0.48
20
0.53
34
2.91
98
0.68
83
2.97
96
1.08
89
3.89
63
1.54
75
0.00
1
0.04
92
0.00
1
0.02
90
0.05
43
0.03
53
CSANtwo views1.05
75
0.61
76
2.46
61
0.09
87
0.08
91
0.71
41
0.71
93
2.20
93
0.93
92
0.85
80
2.92
92
0.79
61
1.46
94
5.37
81
1.58
76
0.01
85
0.01
82
0.00
1
0.00
1
0.09
63
0.05
69
NVStereoNet_ROBtwo views1.08
76
0.39
44
2.15
56
0.08
86
0.02
63
0.45
18
0.02
61
0.89
41
0.67
82
0.10
62
5.85
96
1.90
88
1.11
90
6.20
98
1.14
67
0.20
105
0.00
1
0.03
98
0.07
98
0.12
74
0.13
95
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
MFN_U_SF_DS_RVCtwo views1.16
77
2.03
108
4.57
94
0.02
63
0.04
80
3.52
102
0.27
78
1.80
84
0.58
51
0.22
69
0.28
75
1.38
81
0.02
14
6.08
95
0.42
51
0.14
102
0.68
110
0.01
85
0.26
106
0.39
103
0.47
105
G-Nettwo views1.20
78
0.28
24
2.96
74
0.05
79
0.49
104
1.73
90
0.48
85
0.23
8
0.35
23
1.35
87
9.14
103
2.33
94
0.81
83
2.40
38
1.15
68
0.01
85
0.03
88
0.02
93
0.09
101
0.07
54
0.05
69
STTStereo_v2two views1.20
78
0.28
24
2.96
74
0.05
79
0.49
104
1.73
90
0.48
85
0.23
8
0.35
23
1.35
87
9.14
103
2.33
94
0.81
83
2.40
38
1.15
68
0.01
85
0.03
88
0.02
93
0.09
101
0.07
54
0.05
69
DispFullNettwo views1.20
78
0.63
77
3.58
86
1.00
107
0.82
107
0.41
13
0.00
1
1.44
71
0.61
60
2.55
95
0.64
82
1.10
73
8.95
108
0.84
14
1.22
71
0.00
1
0.02
85
0.00
1
0.00
1
0.24
94
0.02
41
DPSNettwo views1.25
81
0.28
24
5.34
96
0.28
103
0.07
88
1.50
84
0.00
1
3.53
105
0.69
83
0.03
45
0.00
1
0.45
49
2.10
98
5.67
89
4.61
91
0.02
94
0.04
92
0.00
1
0.02
90
0.23
93
0.09
87
DeepPrunerFtwo views1.33
82
0.59
74
8.62
104
0.02
63
0.09
92
0.90
56
0.00
1
1.34
65
6.72
107
0.22
69
0.01
39
0.21
24
0.14
56
5.64
88
1.99
80
0.00
1
0.00
1
0.01
85
0.02
90
0.14
79
0.02
41
ADCP+two views1.36
83
0.76
91
6.73
99
0.01
54
0.02
63
1.32
82
0.80
97
0.81
36
0.72
85
0.01
26
0.01
39
0.96
66
1.05
88
0.33
12
13.50
108
0.00
1
0.00
1
0.00
1
0.00
1
0.06
49
0.02
41
ADCMidtwo views1.38
84
0.90
97
6.15
98
0.04
75
0.03
74
1.67
89
0.18
75
1.13
57
0.93
92
0.89
81
0.58
80
0.86
63
0.71
81
1.03
16
11.96
104
0.01
85
0.00
1
0.29
110
0.08
100
0.17
87
0.03
53
AnyNet_C32two views1.44
85
1.47
103
6.09
97
0.12
91
0.19
99
1.74
92
1.11
100
0.84
39
0.56
38
1.30
86
0.46
79
0.21
24
0.59
77
2.74
45
11.09
102
0.02
94
0.00
1
0.06
102
0.00
1
0.16
84
0.05
69
ADCPNettwo views1.44
85
0.55
70
7.88
102
0.01
54
0.19
99
2.78
97
0.27
78
0.90
42
0.38
25
0.04
51
0.02
58
1.43
83
1.81
97
2.22
34
9.08
98
0.00
1
0.47
108
0.00
1
0.64
111
0.04
39
0.15
98
ADCLtwo views1.52
87
0.44
49
4.70
95
0.07
85
0.03
74
2.21
94
2.35
104
0.92
43
0.80
89
1.57
90
1.11
88
1.25
75
1.55
95
1.77
25
11.53
103
0.01
85
0.00
1
0.00
1
0.00
1
0.10
67
0.01
27
MDST_ROBtwo views1.54
88
0.06
4
3.42
85
0.14
92
0.16
94
5.42
106
0.32
81
4.33
108
0.61
60
6.08
102
1.25
89
1.04
70
1.16
92
5.81
92
0.83
63
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.11
91
SGM_RVCbinarytwo views1.54
88
0.35
37
1.60
31
0.15
93
0.17
96
1.25
81
0.77
94
1.97
88
0.88
91
4.31
99
2.46
91
4.36
100
2.37
99
6.98
103
1.78
77
0.21
106
0.23
105
0.20
108
0.23
105
0.26
95
0.31
102
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
ADCStwo views1.72
90
1.22
102
8.11
103
0.00
1
0.01
55
0.88
53
0.68
91
1.44
71
0.63
77
2.60
96
0.11
67
1.42
82
0.26
68
1.97
27
14.92
110
0.00
1
0.00
1
0.00
1
0.00
1
0.08
60
0.07
82
MSC_U_SF_DS_RVCtwo views1.79
91
1.81
107
3.81
89
0.01
54
0.02
63
12.12
109
0.50
88
2.74
99
0.71
84
0.07
57
0.20
73
2.17
92
0.15
58
6.82
102
1.19
70
0.74
111
0.80
111
0.20
108
0.27
107
0.73
108
0.70
108
SuperBtwo views1.85
92
1.06
101
10.54
106
0.02
63
0.01
55
0.90
56
0.78
95
1.75
83
0.64
80
0.05
53
0.05
62
0.34
40
0.89
86
5.29
80
5.80
96
0.01
85
0.00
1
0.02
93
0.02
90
8.71
113
0.06
77
SANettwo views1.99
93
0.71
85
3.37
83
0.03
70
0.03
74
2.85
98
0.85
98
2.42
96
8.33
109
1.55
89
4.13
94
4.29
99
3.23
101
5.39
82
2.47
84
0.00
1
0.00
1
0.00
1
0.00
1
0.06
49
0.06
77
WCMA_ROBtwo views2.01
94
0.34
32
1.89
43
0.04
75
0.00
1
1.07
71
0.60
89
1.09
53
1.21
96
5.12
100
9.87
107
8.09
104
4.48
103
3.74
61
2.55
85
0.00
1
0.00
1
0.01
85
0.00
1
0.02
22
0.13
95
MSMD_ROBtwo views2.04
95
0.40
46
2.13
55
0.00
1
0.00
1
4.88
105
0.22
76
0.93
44
1.82
101
5.71
101
7.79
101
7.39
103
3.72
102
4.62
72
1.22
71
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.02
41
pmcnntwo views2.10
96
0.59
74
4.52
93
0.27
102
0.86
108
0.34
11
0.29
80
1.69
80
0.75
87
0.36
75
7.50
100
14.39
110
0.56
76
4.90
74
4.69
92
0.00
1
0.00
1
0.00
1
0.00
1
0.16
84
0.04
64
AnyNet_C01two views2.33
97
2.14
110
18.59
109
0.15
93
0.18
97
2.43
96
2.34
103
1.29
62
1.24
97
1.03
82
0.83
87
1.25
75
0.78
82
3.75
62
10.06
99
0.05
98
0.00
1
0.07
104
0.00
1
0.30
100
0.12
94
SGM+DAISYtwo views2.38
98
0.98
99
4.37
92
0.48
104
0.33
103
2.24
95
1.94
102
1.16
58
1.47
99
6.62
103
5.89
97
5.77
101
5.09
104
5.76
91
4.39
90
0.16
103
0.15
104
0.06
102
0.14
103
0.26
95
0.40
104
PWCKtwo views2.44
99
1.65
104
7.31
101
0.22
100
0.02
63
3.12
99
4.74
108
3.06
102
3.48
104
0.72
78
4.42
95
3.40
97
1.67
96
9.06
105
3.36
88
0.39
110
0.01
82
0.00
1
0.00
1
2.19
110
0.08
83
MeshStereopermissivetwo views2.61
100
0.65
78
2.71
70
0.02
63
0.02
63
1.55
86
0.94
99
3.02
101
1.29
98
11.21
106
7.34
99
11.54
107
2.74
100
6.34
100
2.43
83
0.00
1
0.00
1
0.00
1
0.00
1
0.28
97
0.03
53
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
ELAScopylefttwo views2.75
101
0.35
37
2.67
68
0.05
79
0.04
80
6.72
107
1.83
101
2.84
100
3.38
103
8.25
105
9.14
103
3.95
98
6.15
105
3.73
60
5.70
95
0.00
1
0.04
92
0.01
85
0.01
82
0.10
67
0.08
83
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
ELAS_RVCcopylefttwo views2.84
102
0.34
32
3.86
91
0.05
79
0.05
83
1.52
85
5.02
109
1.93
87
5.28
105
7.10
104
6.96
98
6.26
102
8.34
106
4.96
75
4.90
94
0.00
1
0.04
92
0.01
85
0.01
82
0.10
67
0.09
87
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
FC-DCNNcopylefttwo views3.42
103
0.11
12
2.03
48
0.00
1
0.00
1
3.18
100
0.60
89
3.13
103
2.98
102
11.40
107
12.28
108
13.93
108
8.68
107
5.70
90
4.27
89
0.00
1
0.00
1
0.00
1
0.00
1
0.01
11
0.06
77
RTStwo views4.10
104
1.68
105
43.40
111
0.09
87
0.02
63
3.95
103
0.00
1
1.61
76
0.61
60
1.09
83
0.01
39
1.97
89
0.22
65
16.87
107
10.15
100
0.00
1
0.09
100
0.00
1
0.00
1
0.20
89
0.05
69
RTSAtwo views4.10
104
1.68
105
43.40
111
0.09
87
0.02
63
3.95
103
0.00
1
1.61
76
0.61
60
1.09
83
0.01
39
1.97
89
0.22
65
16.87
107
10.15
100
0.00
1
0.09
100
0.00
1
0.00
1
0.20
89
0.05
69
LSMtwo views4.58
106
0.88
96
17.79
108
3.73
112
38.79
113
0.85
51
2.37
105
0.54
22
1.64
100
2.88
97
3.09
93
0.62
55
0.60
78
3.16
51
1.06
65
0.00
1
0.00
1
0.00
1
0.00
1
0.12
74
13.55
113
MADNet+two views4.65
107
2.70
111
46.68
113
0.01
54
0.02
63
2.06
93
0.03
65
3.17
104
1.07
94
0.02
37
0.01
39
1.84
87
0.31
71
21.62
111
13.04
106
0.00
1
0.03
88
0.00
1
0.00
1
0.20
89
0.19
100
DPSimNet_ROBtwo views7.90
108
5.46
112
13.55
107
3.66
111
5.16
112
7.26
108
9.60
112
6.11
109
7.46
108
13.33
108
9.50
106
11.03
105
11.89
109
10.08
106
12.35
105
4.36
112
5.55
112
5.35
113
5.34
113
4.96
111
5.98
112
SGM-ForestMtwo views8.89
109
0.68
81
3.61
87
0.57
106
0.18
97
14.89
112
5.99
110
9.62
111
10.59
110
20.43
110
17.16
110
39.22
112
20.29
111
20.63
110
13.19
107
0.04
96
0.14
103
0.00
1
0.07
98
0.15
81
0.36
103
MADNet++two views9.87
110
2.12
109
7.11
100
7.99
113
4.58
111
14.38
111
6.90
111
9.23
110
5.88
106
16.88
109
17.47
111
11.07
106
15.72
110
24.26
112
19.76
112
8.52
113
6.39
113
4.85
112
2.69
112
5.96
112
5.73
111
MANEtwo views9.96
111
0.69
82
2.88
73
1.22
109
0.61
106
13.15
110
3.46
106
15.62
113
22.27
112
22.74
111
30.00
113
29.74
111
23.22
112
17.17
109
13.78
109
0.29
109
0.56
109
0.09
106
0.61
110
0.15
81
1.00
110
LE_ROBtwo views9.96
111
0.54
66
9.62
105
1.40
110
1.24
110
3.29
101
4.64
107
4.31
107
43.01
113
51.54
113
13.34
109
14.15
109
24.60
114
6.66
101
20.28
113
0.04
96
0.05
98
0.12
107
0.20
104
0.07
54
0.09
87
BEATNet-Init1two views14.52
113
6.56
113
34.32
110
1.10
108
1.02
109
31.53
113
11.99
113
15.58
112
17.42
111
30.17
112
27.03
112
43.14
113
24.36
113
24.41
113
18.27
111
0.27
108
0.32
107
0.36
111
0.46
109
1.26
109
0.76
109
DGTPSM_ROBtwo views81.49
114
60.74
114
99.95
116
61.20
114
86.99
116
99.11
116
100.00
116
95.92
115
99.99
116
73.08
114
100.00
115
70.81
114
100.00
116
77.39
114
99.98
116
35.10
114
97.48
117
35.12
114
99.89
117
46.16
114
90.85
114
DPSMNet_ROBtwo views81.54
115
60.77
115
99.95
116
61.42
115
87.05
117
99.11
116
100.00
116
95.93
116
99.99
116
73.11
115
100.00
115
70.84
115
100.00
116
77.43
115
99.98
116
35.39
115
97.54
118
35.21
115
99.89
117
46.30
115
90.94
115
MEDIAN_ROBtwo views93.37
116
98.75
118
96.05
115
90.90
118
90.53
118
85.63
114
76.45
114
92.03
114
89.53
114
95.05
116
94.99
114
96.02
116
93.85
115
90.36
116
87.21
114
98.36
119
95.79
114
99.61
118
98.77
116
98.80
119
98.72
119
DPSMtwo views94.00
117
79.32
116
100.00
118
83.19
116
85.69
114
100.00
118
100.00
116
100.00
117
100.00
118
100.00
117
100.00
115
100.00
117
100.00
116
100.00
118
100.00
118
89.53
116
97.42
115
74.97
116
95.13
114
83.79
116
90.98
116
DPSM_ROBtwo views94.00
117
79.32
116
100.00
118
83.19
116
85.69
114
100.00
118
100.00
116
100.00
117
100.00
118
100.00
117
100.00
115
100.00
117
100.00
116
100.00
118
100.00
118
89.53
116
97.42
115
74.97
116
95.13
114
83.79
116
90.98
116
AVERAGE_ROBtwo views98.45
119
99.13
119
95.46
114
100.00
120
100.00
120
93.36
115
81.95
115
100.00
117
99.97
115
100.00
117
100.00
115
100.00
117
100.00
116
99.72
117
99.40
115
100.00
120
100.00
120
100.00
119
100.00
119
100.00
120
100.00
120
LSM0two views99.11
120
99.97
120
100.00
118
95.93
119
98.59
119
100.00
118
100.00
116
100.00
117
100.00
118
100.00
117
100.00
115
100.00
117
100.00
116
100.00
118
100.00
118
93.89
118
99.97
119
100.00
119
100.00
119
96.64
118
97.25
118
MSMDNettwo views0.34
72