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
sort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysorted bysort bysort bysort bysort bysort by
DLCB_ROBtwo views4.51
5
0.91
7
3.78
6
2.19
15
1.07
14
6.28
4
3.09
8
9.78
2
7.72
3
10.65
7
12.97
6
13.91
11
3.71
4
8.72
7
5.30
7
0.00
1
0.00
1
0.00
1
0.00
1
0.03
2
0.10
5
CBMV_ROBtwo views4.14
3
0.52
4
3.14
3
1.30
3
0.77
8
6.92
7
1.97
7
10.11
3
9.58
6
8.92
5
14.20
10
7.12
5
5.90
12
8.65
6
3.50
4
0.01
2
0.05
12
0.00
1
0.00
1
0.04
3
0.09
3
iResNet_ROBtwo views4.23
4
1.02
9
4.90
9
2.18
14
0.93
11
2.92
1
0.37
3
15.10
16
16.91
20
7.89
4
10.51
4
7.03
3
3.07
3
8.16
4
3.46
3
0.01
2
0.00
1
0.00
1
0.00
1
0.10
8
0.02
1
NOSS_ROBtwo views3.30
2
0.46
3
2.62
1
2.08
13
1.01
13
5.60
3
0.74
4
10.37
4
11.48
8
5.15
1
8.43
1
5.67
2
1.73
2
7.97
2
2.34
2
0.02
4
0.06
14
0.00
1
0.00
1
0.07
6
0.14
8
XPNet_ROBtwo views6.03
11
1.22
12
5.61
11
2.56
20
0.90
10
6.32
5
7.07
20
12.92
10
8.30
5
14.76
16
15.13
14
19.84
15
6.66
14
10.36
11
8.58
15
0.02
4
0.04
10
0.00
1
0.03
18
0.11
10
0.24
15
MDST_ROBtwo views8.37
17
0.32
1
9.03
20
4.18
23
2.42
21
26.86
29
6.14
13
19.36
24
13.52
11
27.09
24
22.75
17
9.47
6
4.74
7
15.06
21
6.34
8
0.02
4
0.02
7
0.00
1
0.00
1
0.02
1
0.13
7
PSMNet_ROBtwo views5.02
7
1.63
19
6.03
12
1.90
11
1.83
19
9.57
10
6.35
17
15.58
17
7.23
2
6.15
3
10.48
3
12.22
10
4.16
5
8.02
3
8.71
16
0.02
4
0.01
5
0.01
15
0.10
21
0.20
14
0.12
6
CBMVpermissivetwo views5.35
8
0.91
7
3.67
5
1.62
7
0.44
3
10.09
13
7.19
22
12.49
7
12.33
10
12.22
10
14.69
11
10.93
7
6.48
13
8.51
5
4.96
6
0.02
4
0.15
22
0.00
1
0.00
1
0.17
13
0.17
11
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
DN-CSS_ROBtwo views2.69
1
1.40
16
5.34
10
2.31
17
0.75
7
3.14
2
0.06
1
6.11
1
3.87
1
5.34
2
12.18
5
2.34
1
1.22
1
7.84
1
1.48
1
0.03
9
0.00
1
0.00
1
0.00
1
0.35
19
0.03
2
LALA_ROBtwo views6.58
12
1.80
21
6.25
13
1.26
2
0.94
12
10.08
12
9.02
23
16.00
18
11.51
9
12.74
12
13.02
7
24.77
19
5.25
9
10.56
12
8.02
12
0.04
10
0.05
12
0.00
1
0.02
15
0.10
8
0.25
16
ETE_ROBtwo views5.80
9
1.77
20
6.33
15
1.44
5
0.78
9
6.43
6
6.90
19
12.53
8
8.08
4
12.93
13
14.89
12
21.13
18
5.87
11
9.83
10
6.57
10
0.04
10
0.01
5
0.00
1
0.02
15
0.08
7
0.33
18
NaN_ROBtwo views6.00
10
1.24
13
6.29
14
1.34
4
1.68
18
9.60
11
10.31
27
15.09
15
15.79
17
12.62
11
8.95
2
11.67
9
5.83
10
11.78
13
6.41
9
0.05
12
0.13
21
0.08
21
0.20
22
0.22
15
0.79
22
LE_ROBtwo views16.73
30
1.28
15
11.61
26
3.72
22
1.65
17
16.67
23
9.17
24
14.39
12
55.91
32
63.81
32
40.86
30
35.94
28
37.73
32
14.24
20
26.87
31
0.05
12
0.10
20
0.13
22
0.22
23
0.12
11
0.15
9
pmcnntwo views7.72
14
1.27
14
9.42
22
2.91
21
3.14
22
9.44
9
6.23
14
12.56
9
16.51
19
14.53
15
24.08
18
27.44
23
8.49
15
9.32
9
8.44
14
0.06
14
0.08
19
0.00
1
0.00
1
0.30
16
0.15
9
SANettwo views10.64
23
1.86
22
10.91
24
1.76
8
0.71
6
14.62
21
9.23
25
19.18
22
37.14
30
19.22
19
27.96
20
25.86
21
19.11
23
13.02
17
10.63
21
0.08
15
0.06
14
0.03
18
0.02
15
0.62
22
0.81
23
CSANtwo views7.62
13
1.60
18
6.56
16
1.83
9
0.66
5
12.40
16
10.52
29
14.45
13
21.32
23
14.19
14
15.98
15
17.84
13
13.02
19
12.32
14
8.38
13
0.09
16
0.07
17
0.03
18
0.04
19
0.33
17
0.67
21
SGM-Foresttwo views4.96
6
0.32
1
2.84
2
1.21
1
0.64
4
10.23
14
6.64
18
11.55
5
10.98
7
10.94
8
13.59
8
11.65
8
4.30
6
8.94
8
4.63
5
0.11
17
0.04
10
0.00
1
0.00
1
0.05
4
0.46
20
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
MeshStereopermissivetwo views11.52
24
1.52
17
4.55
7
1.89
10
1.46
16
19.87
25
5.11
11
20.66
26
15.91
18
32.67
28
34.51
26
39.34
31
21.15
25
18.74
24
12.10
23
0.11
17
0.06
14
0.01
15
0.00
1
0.45
20
0.22
13
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
SGM_ROBbinarytwo views10.08
22
0.60
5
3.42
4
2.30
16
0.32
1
19.41
24
6.33
16
18.95
21
14.64
13
25.14
23
24.32
19
33.34
26
18.79
22
19.86
25
12.55
24
0.25
19
0.26
24
0.22
24
0.24
24
0.34
18
0.40
19
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
MSMD_ROBtwo views9.28
20
1.09
11
4.65
8
1.58
6
0.39
2
16.52
22
4.41
10
13.60
11
14.87
15
22.34
21
39.89
28
25.67
20
20.71
24
12.42
15
6.98
11
0.34
20
0.03
9
0.00
1
0.00
1
0.05
4
0.09
3
PWC_ROBbinarytwo views8.24
16
3.13
26
12.74
27
2.43
18
4.43
24
7.51
8
1.22
5
16.63
20
19.24
22
16.08
17
28.29
21
13.99
12
10.16
18
13.63
19
14.06
25
0.42
21
0.00
1
0.05
20
0.00
1
0.59
21
0.27
17
FBW_ROBtwo views8.50
18
1.03
10
7.98
19
1.93
12
1.28
15
13.10
18
6.23
14
22.50
27
18.98
21
18.82
18
14.91
13
19.06
14
10.04
17
18.41
23
9.83
19
0.62
22
0.22
23
1.82
27
0.82
26
0.99
23
1.36
24
WCMA_ROBtwo views9.21
19
0.87
6
7.37
17
2.54
19
2.13
20
13.59
19
5.80
12
11.64
6
14.01
12
24.43
22
32.99
24
27.09
22
18.02
20
12.51
16
9.85
20
0.81
23
0.07
17
0.01
15
0.01
14
0.16
12
0.23
14
ELAScopylefttwo views16.72
29
2.14
24
9.23
21
4.92
24
4.53
26
32.66
31
15.11
30
27.40
31
28.68
28
40.27
30
44.90
31
38.33
30
30.50
31
26.44
30
21.94
30
0.88
24
1.23
27
0.67
25
0.89
27
1.49
26
2.18
26
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
ELAS_ROBcopylefttwo views16.54
28
2.26
25
10.09
23
5.50
25
4.46
25
28.28
30
16.72
31
25.55
30
33.54
29
40.19
29
40.30
29
36.68
29
30.03
30
29.40
31
20.61
28
0.98
25
1.21
26
0.86
26
0.70
25
1.39
25
2.16
25
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
PWCDC_ROBbinarytwo views7.92
15
3.17
27
7.48
18
5.73
26
4.40
23
10.45
15
0.35
2
14.52
14
28.19
26
10.36
6
31.27
23
7.04
4
9.14
16
13.22
18
8.78
17
2.74
26
0.02
7
0.00
1
0.00
1
1.31
24
0.17
11
DispFullNettwo views17.47
31
26.01
31
33.98
31
22.58
31
20.86
32
13.84
20
1.28
6
16.50
19
26.27
25
19.97
20
17.17
16
20.52
17
18.49
21
22.86
27
10.76
22
5.13
27
2.83
28
30.72
32
7.72
31
20.86
31
11.01
32
PDISCO_ROBtwo views9.62
21
1.99
23
11.51
25
9.88
29
9.61
29
21.48
26
3.83
9
19.33
23
28.49
27
11.27
9
14.17
9
19.92
16
5.02
8
16.35
22
9.18
18
5.28
28
0.41
25
0.14
23
0.09
20
2.05
27
2.36
27
NVStereoNet_ROBtwo views16.04
27
6.75
29
12.90
28
6.37
27
7.42
28
12.89
17
9.74
26
22.78
28
25.12
24
30.32
25
46.19
32
34.37
27
25.38
28
21.48
26
21.38
29
5.94
29
3.10
29
6.07
30
10.09
32
4.01
28
8.54
31
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
SPS-STEREOcopylefttwo views15.04
25
6.23
28
13.21
29
11.34
30
11.65
31
23.30
27
7.15
21
24.16
29
15.65
16
31.78
27
29.19
22
31.62
24
21.32
26
24.62
28
19.50
27
7.59
30
4.19
31
3.22
28
1.48
28
6.99
30
6.54
28
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
SGM+DAISYtwo views15.62
26
7.26
30
19.28
30
8.94
28
10.11
30
26.25
28
10.49
28
19.36
24
14.65
14
30.64
26
33.59
25
33.00
25
22.32
27
24.96
29
16.42
26
7.90
31
6.25
32
4.51
29
3.37
29
5.86
29
7.20
29
PWCKtwo views30.53
32
44.32
32
47.25
33
29.76
32
7.23
27
40.78
32
27.10
32
44.73
32
44.32
31
47.31
31
36.37
27
47.16
32
26.05
29
41.26
32
31.87
32
21.83
32
4.03
30
29.50
31
4.67
30
27.17
32
7.80
30
DPSimNet_ROBtwo views53.45
33
64.73
33
44.39
32
53.97
33
45.39
33
53.66
33
54.83
33
55.15
33
57.87
33
64.16
33
50.83
33
63.40
33
53.34
33
46.45
33
65.81
33
63.13
33
26.54
33
57.94
33
51.11
33
45.52
33
50.69
33
DPSMtwo views99.95
38
100.00
36
100.00
38
99.76
35
100.00
35
100.00
36
100.00
36
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
100.00
38
99.21
34
100.00
35
100.00
34
100.00
35
99.99
35
99.95
35
DPSM_ROBtwo views99.95
38
100.00
36
100.00
38
99.76
35
100.00
35
100.00
36
100.00
36
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
100.00
38
99.21
34
100.00
35
100.00
34
100.00
35
99.99
35
99.95
35
MEDIAN_ROBtwo views98.41
34
99.70
34
99.30
35
97.09
34
97.02
34
96.89
34
95.77
35
97.66
34
97.28
34
98.79
36
98.94
34
99.18
34
98.14
34
96.89
34
96.88
34
99.96
36
99.16
34
100.00
34
99.99
34
99.69
34
99.88
34
DGTPSM_ROBtwo views99.90
36
100.00
36
99.99
36
99.99
37
100.00
35
100.00
36
100.00
36
99.97
35
100.00
35
98.35
34
100.00
35
99.84
35
100.00
35
99.98
35
99.99
35
99.99
37
100.00
35
100.00
34
100.00
35
100.00
37
100.00
37
AVERAGE_ROBtwo views99.62
35
99.95
35
98.81
34
100.00
39
100.00
35
98.08
35
95.47
34
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
100.00
35
100.00
37
99.99
35
100.00
38
100.00
35
100.00
34
100.00
35
100.00
37
100.00
37
DPSMNet_ROBtwo views99.91
37
100.00
36
99.99
36
99.99
37
100.00
35
100.00
36
100.00
36
99.98
36
100.00
35
98.35
34
100.00
35
99.84
35
100.00
35
99.98
35
99.99
35
100.00
38
100.00
35
100.00
34
100.00
35
100.00
37
100.00
37