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 bysorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
Anonymous3two views1.76
14
0.44
18
6.37
101
0.33
12
0.50
40
4.33
59
0.54
42
4.43
10
2.48
1
2.21
8
3.24
10
1.53
19
1.14
17
5.64
41
1.29
26
0.17
119
0.23
142
0.00
1
0.00
1
0.12
64
0.15
79
sCroCo_RVCtwo views1.21
3
0.65
47
5.16
78
1.09
32
1.29
74
0.92
1
0.19
22
3.92
6
2.68
2
0.30
1
3.04
9
0.52
11
1.56
26
1.58
8
0.78
12
0.09
106
0.25
146
0.00
1
0.00
1
0.07
45
0.11
67
XX-Stereotwo views2.71
38
0.38
13
5.96
94
2.25
97
2.20
106
1.54
6
0.21
23
5.44
22
3.05
3
2.28
10
16.16
120
7.03
60
0.24
1
6.19
45
0.79
13
0.00
1
0.15
129
0.00
1
0.23
135
0.13
72
0.04
38
R-Stereo Traintwo views2.44
30
0.32
2
1.93
2
0.94
26
0.16
17
3.67
47
0.61
49
6.37
31
3.08
4
9.14
95
17.44
127
1.80
21
0.77
11
1.76
10
0.70
8
0.00
1
0.01
54
0.00
1
0.00
1
0.01
7
0.03
27
RAFT-Stereopermissivetwo views2.44
30
0.32
2
1.93
2
0.94
26
0.16
17
3.67
47
0.61
49
6.37
31
3.08
4
9.14
95
17.44
127
1.80
21
0.77
11
1.76
10
0.70
8
0.00
1
0.01
54
0.00
1
0.00
1
0.01
7
0.03
27
Lahav Lipson, Zachary Teed, and Jia Deng: RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching. 3DV
GEStwo views2.49
33
0.38
13
4.29
59
1.58
53
0.05
3
4.58
68
1.07
69
9.70
59
3.39
6
5.21
34
6.64
36
1.84
23
3.36
50
4.86
37
2.54
45
0.06
100
0.01
54
0.06
127
0.01
82
0.12
64
0.03
27
sAnonymous2two views1.35
4
1.66
125
8.14
124
0.55
17
0.46
37
1.23
2
0.24
26
4.00
7
3.47
7
0.40
2
1.11
1
0.23
2
1.42
23
0.96
4
0.57
6
0.01
48
0.00
1
0.00
1
0.20
131
2.07
163
0.22
95
CroCo_RVCtwo views1.35
4
1.66
125
8.14
124
0.55
17
0.46
37
1.23
2
0.24
26
4.00
7
3.47
7
0.40
2
1.11
1
0.23
2
1.42
23
0.96
4
0.57
6
0.01
48
0.00
1
0.00
1
0.20
131
2.07
163
0.22
95
Anonymoustwo views1.37
6
0.52
25
4.23
57
1.38
41
2.70
121
1.85
17
0.42
36
4.77
12
3.54
9
0.53
4
1.56
5
0.19
1
1.32
20
3.11
23
0.92
19
0.05
96
0.08
110
0.00
1
0.00
1
0.17
83
0.15
79
MSMDNettwo views2.02
20
0.37
11
2.84
26
1.14
34
0.34
30
3.73
50
0.53
40
6.21
29
3.55
10
5.32
37
5.17
23
5.85
52
1.74
30
2.23
18
0.91
16
0.01
48
0.10
116
0.00
1
0.24
137
0.02
14
0.02
14
iRaftStereo_RVCtwo views1.62
11
0.68
49
1.98
5
0.67
21
0.05
3
2.99
37
0.04
8
5.36
18
3.74
11
3.82
21
4.96
18
2.54
28
0.62
6
3.35
25
0.97
20
0.00
1
0.07
107
0.00
1
0.00
1
0.30
108
0.27
111
DN-CSS_ROBtwo views2.69
37
1.40
110
5.34
81
2.31
102
0.75
49
3.14
39
0.06
9
6.11
28
3.87
12
5.34
38
12.18
90
2.34
27
1.22
18
7.84
68
1.48
28
0.03
79
0.00
1
0.00
1
0.00
1
0.35
117
0.03
27
TestStereotwo views1.74
13
1.02
84
3.64
44
0.20
3
0.28
26
1.33
4
0.01
1
5.24
17
4.01
13
3.18
13
4.97
19
2.07
25
0.70
8
7.32
59
0.75
11
0.02
65
0.00
1
0.00
1
0.00
1
0.01
7
0.00
1
raftrobusttwo views1.86
17
0.74
54
1.94
4
2.15
89
2.24
108
2.74
28
0.53
40
4.89
13
4.05
14
6.09
49
3.89
12
1.64
20
2.55
39
2.44
21
1.12
24
0.00
1
0.03
81
0.00
1
0.00
1
0.12
64
0.08
52
DIP-Stereotwo views1.97
19
0.64
44
2.95
28
0.17
2
0.10
10
4.83
71
0.13
15
8.60
47
4.06
15
6.42
58
4.92
17
0.44
8
0.72
10
3.57
26
1.80
32
0.00
1
0.01
54
0.00
1
0.00
1
0.05
39
0.04
38
Zihua Zheng, Ni Nie, Zhi Ling, Pengfei Xiong, Jiangyu Liu, Hao Wang, Jiankun Li: DIP: Deep Inverse Patchmatch for High-Resolution Optical Flow. cvpr2022
TANstereotwo views1.20
2
0.49
23
2.04
6
0.34
13
0.10
10
1.80
16
0.01
1
3.40
4
4.10
16
4.37
23
4.82
16
0.41
6
0.70
8
0.94
3
0.46
4
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
EAI-Stereotwo views2.31
29
0.44
18
2.72
22
0.68
22
0.56
41
3.51
44
2.62
90
11.42
89
4.14
17
2.09
7
7.23
43
1.21
14
0.91
14
5.67
42
1.51
29
0.01
48
0.25
146
0.00
1
0.88
163
0.04
31
0.26
108
ac_64two views3.70
62
0.47
21
3.50
40
2.87
125
3.50
137
5.91
82
0.72
60
10.17
69
4.34
18
8.16
85
5.80
31
10.82
100
5.64
96
7.08
56
4.73
80
0.03
79
0.00
1
0.00
1
0.00
1
0.08
48
0.14
76
RALCasStereoNettwo views1.93
18
1.01
82
2.75
23
1.88
67
1.75
91
2.88
34
0.18
21
4.64
11
4.36
19
2.27
9
5.14
22
2.22
26
0.79
13
7.10
57
1.00
22
0.00
1
0.34
152
0.00
1
0.00
1
0.10
53
0.16
83
DeepPruner_ROBtwo views3.52
57
1.14
96
4.06
55
1.12
33
1.65
84
3.65
46
0.83
64
13.96
121
4.47
20
7.80
79
10.84
76
7.05
64
2.16
35
8.14
77
3.08
55
0.07
104
0.03
81
0.00
1
0.01
82
0.32
112
0.06
45
raft+_RVCtwo views1.79
15
0.84
64
2.32
12
2.40
107
0.85
56
1.38
5
0.09
11
6.36
30
4.75
21
2.85
11
5.46
30
1.17
13
1.36
22
4.54
33
0.91
16
0.00
1
0.10
116
0.00
1
0.00
1
0.11
59
0.22
95
DMCA-RVCcopylefttwo views3.17
46
1.26
103
6.39
103
2.10
85
1.51
80
3.02
38
0.56
46
7.84
41
5.06
22
9.23
98
5.19
27
8.46
79
2.72
42
6.31
48
3.22
60
0.06
100
0.11
120
0.09
135
0.02
89
0.16
81
0.08
52
DMCAtwo views2.82
41
0.64
44
4.45
63
1.61
56
0.89
58
3.86
53
0.57
47
9.18
52
5.10
23
6.24
54
6.49
34
7.17
67
2.13
34
3.31
24
4.73
80
0.00
1
0.01
54
0.00
1
0.00
1
0.04
31
0.02
14
HITNettwo views2.79
40
0.77
55
4.02
54
2.03
80
0.11
13
5.58
78
0.59
48
9.24
54
5.15
24
6.42
58
7.26
44
3.66
34
2.92
46
4.07
31
3.87
71
0.00
1
0.00
1
0.00
1
0.00
1
0.06
43
0.02
14
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
UNettwo views3.97
72
0.61
38
6.50
106
1.68
61
3.22
127
9.82
117
0.31
29
9.50
56
5.32
25
7.20
70
6.35
33
10.59
97
5.91
100
6.93
55
5.49
87
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.01
8
PMTNettwo views1.41
7
0.37
11
2.06
8
0.26
8
0.15
16
2.94
36
0.34
31
3.39
3
5.33
26
1.00
6
1.27
3
0.28
5
0.93
15
4.04
30
0.74
10
5.00
173
0.00
1
0.00
1
0.00
1
0.03
26
0.01
8
HCRNettwo views2.56
34
0.53
29
3.35
36
0.24
7
3.39
134
2.19
18
0.23
25
6.89
35
5.50
27
5.26
35
4.79
15
5.27
43
2.78
44
7.71
64
2.89
50
0.00
1
0.02
70
0.00
1
0.00
1
0.04
31
0.04
38
BEATNet_4xtwo views3.24
48
1.27
104
5.89
92
1.56
52
0.10
10
5.26
75
1.07
69
10.08
66
5.50
27
6.89
66
7.73
51
4.53
38
4.13
64
5.05
39
5.27
84
0.04
88
0.05
93
0.00
1
0.00
1
0.23
103
0.23
99
RALAANettwo views2.46
32
1.09
91
3.70
46
2.55
116
0.80
54
3.37
41
0.15
18
7.44
39
5.54
29
5.65
41
6.01
32
3.54
33
1.13
16
6.34
49
1.58
30
0.00
1
0.01
54
0.00
1
0.05
110
0.21
95
0.03
27
FENettwo views2.19
25
0.34
9
3.04
31
2.05
82
0.26
25
2.82
31
0.13
15
7.02
36
5.55
30
4.97
30
1.94
6
4.91
39
3.39
51
4.71
35
2.50
43
0.02
65
0.03
81
0.00
1
0.00
1
0.00
1
0.21
93
acv_fttwo views3.69
61
0.53
29
4.58
68
2.31
102
4.71
162
7.44
101
0.86
65
8.54
46
5.70
31
9.29
99
9.26
64
5.77
49
4.15
65
2.24
19
8.13
115
0.00
1
0.00
1
0.00
1
0.00
1
0.29
107
0.01
8
ACVNettwo views2.58
35
0.53
29
2.69
18
0.84
23
0.59
42
2.61
25
0.55
44
8.35
42
5.70
31
4.45
24
9.26
64
5.77
49
3.50
53
2.24
19
4.41
76
0.00
1
0.00
1
0.00
1
0.00
1
0.17
83
0.01
8
PSMNet-RSSMtwo views3.33
53
0.77
55
3.22
35
2.12
86
1.90
99
2.33
20
0.71
59
11.17
86
5.83
33
5.90
45
11.43
82
7.04
62
4.83
85
5.96
44
3.17
58
0.00
1
0.00
1
0.02
107
0.00
1
0.08
48
0.02
14
RAFT-Stereo + iAFFtwo views2.12
24
0.58
34
2.43
14
1.27
37
0.16
17
1.60
7
0.14
17
5.17
15
5.90
34
7.10
68
10.10
69
3.06
31
1.74
30
1.87
14
0.97
20
0.01
48
0.02
70
0.00
1
0.00
1
0.23
103
0.02
14
Gwc-CoAtRStwo views2.02
20
0.33
7
2.61
16
1.04
29
0.30
27
3.56
45
0.40
34
5.89
26
5.94
35
5.08
31
5.01
21
5.55
47
1.72
28
1.60
9
0.91
16
0.01
48
0.09
115
0.00
1
0.25
139
0.01
7
0.02
14
UPFNettwo views3.82
67
0.47
21
4.34
61
2.21
94
3.38
132
7.26
98
2.69
92
9.89
61
5.96
36
7.98
82
7.53
48
7.89
76
3.15
48
7.72
66
5.88
91
0.01
48
0.02
70
0.00
1
0.00
1
0.04
31
0.06
45
ARAFTtwo views3.13
45
1.11
94
7.34
116
0.42
16
0.14
15
4.52
67
2.99
95
7.41
38
5.97
37
8.85
93
7.06
42
5.20
42
1.25
19
8.49
82
1.44
27
0.00
1
0.03
81
0.00
1
0.00
1
0.33
113
0.11
67
rafts_anoytwo views1.72
12
0.79
59
2.58
15
2.03
80
1.75
91
2.47
23
0.16
20
5.48
23
6.01
38
3.24
15
2.11
7
1.92
24
1.33
21
3.08
22
1.13
25
0.01
48
0.18
136
0.00
1
0.01
82
0.03
26
0.07
47
APVNettwo views5.98
102
1.72
128
8.72
127
3.21
139
4.04
150
12.31
135
1.81
84
19.09
155
6.10
39
6.13
50
13.41
101
10.69
99
6.30
101
11.26
122
13.90
148
0.05
96
0.02
70
0.11
137
0.03
99
0.53
129
0.20
92
GANet-RSSMtwo views3.27
49
0.60
36
3.38
37
2.94
131
2.00
100
2.61
25
0.48
38
10.93
85
6.11
40
5.47
39
10.20
71
6.57
57
5.44
94
6.25
47
2.23
41
0.00
1
0.00
1
0.08
132
0.00
1
0.17
83
0.03
27
pcwnet_v2two views2.93
42
0.29
1
3.01
29
2.39
106
1.07
67
2.73
27
0.89
66
11.37
87
6.25
41
6.20
52
11.49
83
2.71
30
2.74
43
4.22
32
2.75
49
0.06
100
0.05
93
0.00
1
0.00
1
0.01
7
0.37
118
cf-rtwo views3.42
55
0.81
61
3.97
53
2.41
109
2.78
122
2.85
33
0.64
54
8.44
45
6.32
42
6.25
55
11.57
84
8.97
84
3.54
55
5.93
43
3.83
70
0.00
1
0.00
1
0.00
1
0.00
1
0.09
52
0.07
47
psm_uptwo views3.85
69
0.82
62
3.01
29
2.00
78
2.58
119
5.28
76
3.34
102
10.85
83
6.33
43
7.27
72
7.32
45
10.15
91
9.02
124
6.19
45
2.51
44
0.22
126
0.01
54
0.00
1
0.00
1
0.12
64
0.03
27
CREStereotwo views0.98
1
0.42
17
1.86
1
0.07
1
0.22
23
1.65
8
0.02
7
4.07
9
6.33
43
0.72
5
1.54
4
0.43
7
0.64
7
1.22
7
0.44
3
0.00
1
0.00
1
0.00
1
0.00
1
0.01
7
0.00
1
Jiankun Li, Peisen Wang, Pengfei Xiong, Tao Cai, Ziwei Yan, Lei Yang, Jiangyu Liu, Haoqiang Fan, Shuaicheng Liu: Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlation. CVPR 2022
HSMtwo views4.00
74
0.79
59
3.16
34
1.59
55
2.17
103
6.77
94
1.11
71
12.28
104
6.35
45
6.75
62
8.11
54
13.90
117
5.37
90
8.85
90
2.71
47
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.02
14
FADNet_RVCtwo views3.91
71
1.67
127
12.95
147
0.96
28
0.75
49
5.71
81
0.54
42
10.83
82
6.60
46
3.46
16
8.09
53
4.10
36
3.40
52
9.43
104
6.33
95
0.36
140
0.44
158
0.17
145
0.46
156
0.91
146
0.95
150
AFF-stereotwo views2.59
36
0.68
49
2.35
13
1.55
51
0.17
20
1.67
9
0.08
10
5.75
25
6.60
46
6.20
52
12.13
87
6.16
55
2.36
37
4.75
36
1.11
23
0.02
65
0.02
70
0.00
1
0.00
1
0.23
103
0.05
43
STTStereotwo views3.60
58
0.93
76
6.34
100
2.71
121
2.23
107
3.68
49
0.63
52
9.42
55
6.73
48
9.87
106
6.97
40
8.84
83
3.65
57
6.85
53
3.04
53
0.00
1
0.02
70
0.01
95
0.00
1
0.02
14
0.02
14
HGLStereotwo views4.25
83
0.52
25
4.65
69
3.09
134
5.18
164
6.51
92
0.73
61
10.37
72
6.74
49
7.32
73
12.44
95
7.00
59
7.21
111
7.46
61
5.51
88
0.00
1
0.02
70
0.00
1
0.00
1
0.06
43
0.10
64
UCFNet_RVCtwo views3.09
43
0.77
55
2.27
11
2.35
104
1.79
95
3.50
43
1.06
68
12.12
101
6.91
50
4.71
27
7.67
50
7.58
70
4.40
72
4.95
38
1.67
31
0.00
1
0.00
1
0.00
1
0.00
1
0.11
59
0.01
8
CFNet-RSSMtwo views2.11
23
0.33
7
2.71
21
1.04
29
0.30
27
3.21
40
0.41
35
6.67
34
6.96
51
5.62
40
4.58
14
5.40
45
1.71
27
1.86
13
0.87
15
0.00
1
0.08
110
0.00
1
0.32
144
0.02
14
0.02
14
MaskLacGwcNet_RVCtwo views5.97
101
5.36
164
5.02
73
3.75
153
3.57
139
15.81
148
4.27
116
11.72
93
6.96
51
7.23
71
8.47
57
9.09
85
2.66
40
9.36
103
2.71
47
7.49
178
2.32
174
2.01
175
0.08
117
6.77
177
4.67
174
DSFCAtwo views4.21
80
0.50
24
5.45
83
1.34
39
1.68
88
7.04
97
4.51
122
10.73
79
7.00
53
10.78
114
6.80
37
8.48
81
4.63
76
9.91
108
5.12
83
0.02
65
0.06
100
0.02
107
0.03
99
0.08
48
0.03
27
DANettwo views6.02
104
1.23
100
8.45
126
3.86
156
3.94
149
7.64
103
1.34
77
9.51
57
7.00
53
13.39
133
15.53
117
15.99
128
7.02
110
12.14
126
12.37
143
0.19
122
0.12
125
0.02
107
0.03
99
0.13
72
0.56
136
Z Ling, K Yang, J Li, Y Zhang, X Gao, L Luo, L Xie: Domain-adaptive modules for stereo matching network. Neurocomputing 2021
psmorigintwo views7.34
126
1.58
119
18.31
158
2.35
104
0.87
57
6.72
93
1.70
83
10.63
77
7.14
55
19.77
150
22.71
140
20.13
149
11.34
135
12.96
133
9.49
127
0.04
88
0.17
134
0.06
127
0.17
128
0.21
95
0.50
131
FADNet-RVCtwo views3.98
73
1.84
134
12.48
144
1.69
62
0.44
35
4.33
59
1.31
76
11.84
96
7.15
56
3.53
18
3.50
11
10.63
98
4.43
73
9.12
95
6.25
94
0.03
79
0.10
116
0.00
1
0.03
99
0.60
135
0.25
105
PSMNet_ROBtwo views5.02
94
1.63
123
6.03
95
1.90
70
1.83
96
9.57
115
6.35
141
15.58
138
7.23
57
6.15
51
10.48
73
12.22
109
4.16
66
8.02
71
8.71
121
0.02
65
0.01
54
0.01
95
0.10
123
0.20
90
0.12
71
MLCVtwo views3.44
56
0.88
68
5.60
86
1.39
42
0.25
24
4.36
61
0.33
30
7.25
37
7.28
58
9.17
97
12.24
92
5.09
41
2.47
38
9.15
97
3.23
61
0.00
1
0.00
1
0.00
1
0.00
1
0.10
53
0.02
14
G-Nettwo views8.41
138
1.54
117
10.97
137
5.73
169
3.60
140
26.19
167
4.41
120
10.10
67
7.42
59
19.71
149
24.99
148
14.38
121
15.83
150
10.99
118
9.53
128
0.50
151
0.46
160
0.19
148
0.25
139
0.80
143
0.66
141
CFNet-ftpermissivetwo views3.31
51
0.94
78
2.69
18
1.50
48
2.38
111
2.81
29
0.68
55
8.35
42
7.43
60
4.45
24
9.94
67
10.20
92
4.60
74
6.50
52
3.41
64
0.00
1
0.00
1
0.03
115
0.00
1
0.22
98
0.03
27
CFNet_RVCtwo views3.31
51
0.94
78
2.69
18
1.50
48
2.38
111
2.81
29
0.68
55
8.35
42
7.43
60
4.45
24
9.94
67
10.20
92
4.60
74
6.49
51
3.41
64
0.00
1
0.00
1
0.03
115
0.00
1
0.22
98
0.03
27
XX-TBDtwo views1.44
8
0.85
66
2.04
6
2.45
111
0.05
3
2.34
21
0.01
1
3.46
5
7.51
62
2.88
12
4.98
20
0.26
4
0.25
2
1.15
6
0.46
4
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
0.00
1
hitnet-ftcopylefttwo views3.66
59
0.65
47
2.79
25
0.89
25
0.72
48
4.25
58
0.55
44
9.11
51
7.52
63
7.14
69
10.19
70
12.49
111
4.70
78
7.44
60
4.41
76
0.03
79
0.00
1
0.06
127
0.00
1
0.20
90
0.04
38
RAFT + AFFtwo views4.30
86
1.55
118
5.84
90
1.43
45
2.36
110
5.94
83
6.72
145
10.01
63
7.56
64
7.55
76
8.96
63
7.09
65
6.73
109
8.57
85
4.02
73
0.02
65
0.47
161
0.00
1
0.01
82
0.56
130
0.54
135
NLCA_NET_v2_RVCtwo views3.84
68
1.06
88
5.23
79
2.72
122
3.27
128
4.36
61
0.61
49
10.71
78
7.56
64
8.75
90
7.89
52
9.86
89
3.90
62
7.15
58
3.44
66
0.14
113
0.02
70
0.02
107
0.03
99
0.04
31
0.03
27
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
DLCB_ROBtwo views4.51
88
0.91
71
3.78
47
2.19
92
1.07
67
6.28
85
3.09
97
9.78
60
7.72
66
10.65
112
12.97
98
13.91
118
3.71
59
8.72
87
5.30
85
0.00
1
0.00
1
0.00
1
0.00
1
0.03
26
0.10
64
AdaStereotwo views3.09
43
0.58
34
3.04
31
2.84
124
0.48
39
4.08
56
1.29
75
12.16
103
7.77
67
6.03
47
9.62
66
5.79
51
1.53
25
4.56
34
1.93
37
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.02
14
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.
GEStereo_RVCtwo views3.74
64
0.73
53
4.95
72
2.01
79
1.24
71
4.36
61
3.07
96
11.77
95
7.80
68
6.87
65
6.52
35
7.70
73
4.31
71
8.98
93
4.19
74
0.03
79
0.01
54
0.02
107
0.02
89
0.13
72
0.11
67
delettwo views4.27
84
0.72
52
4.50
64
1.07
31
3.75
146
10.79
127
4.04
112
9.95
62
7.89
69
8.09
83
8.89
60
8.46
79
3.81
60
7.56
63
5.80
90
0.00
1
0.00
1
0.00
1
0.00
1
0.11
59
0.02
14
GwcNet-RSSMtwo views3.77
65
1.05
87
5.06
75
2.21
94
1.87
97
2.36
22
1.52
78
8.98
50
7.96
70
5.91
46
14.59
110
7.94
77
3.50
53
6.43
50
5.70
89
0.00
1
0.00
1
0.04
122
0.00
1
0.20
90
0.08
52
RASNettwo views4.52
90
0.61
38
4.42
62
3.42
144
4.68
161
4.58
68
0.99
67
9.54
58
8.01
71
5.28
36
11.42
81
10.34
96
8.88
122
9.28
100
8.68
120
0.15
115
0.00
1
0.00
1
0.00
1
0.03
26
0.04
38
ETE_ROBtwo views5.80
98
1.77
131
6.33
99
1.44
47
0.78
53
6.43
89
6.90
147
12.53
107
8.08
72
12.93
131
14.89
113
21.13
153
5.87
98
9.83
106
6.57
99
0.04
88
0.01
54
0.00
1
0.02
89
0.08
48
0.33
116
CREStereo++_RVCtwo views1.59
9
0.39
15
2.26
9
0.36
14
0.01
1
1.67
9
0.01
1
2.95
1
8.12
73
3.69
19
7.44
46
0.47
9
0.32
3
0.74
1
0.39
1
0.00
1
0.00
1
0.00
1
3.03
175
0.00
1
0.00
1
s12784htwo views1.59
9
0.39
15
2.26
9
0.36
14
0.01
1
1.67
9
0.01
1
2.95
1
8.12
73
3.69
19
7.44
46
0.47
9
0.32
3
0.74
1
0.39
1
0.00
1
0.00
1
0.00
1
3.03
175
0.00
1
0.00
1
GANettwo views6.22
107
1.07
89
4.07
56
2.27
98
0.89
58
9.19
110
9.52
157
12.02
98
8.13
75
10.72
113
29.09
155
13.86
116
7.52
114
11.00
119
4.39
75
0.36
140
0.00
1
0.02
107
0.02
89
0.12
64
0.08
52
PS-NSSStwo views3.90
70
1.84
134
4.55
65
0.87
24
2.95
125
4.10
57
2.00
86
10.77
80
8.23
76
6.75
62
14.63
111
5.01
40
2.24
36
8.41
81
2.91
51
0.33
138
0.03
81
0.83
166
0.00
1
0.82
144
0.70
143
XPNet_ROBtwo views6.03
105
1.22
99
5.61
87
2.56
117
0.90
60
6.32
86
7.07
148
12.92
111
8.30
77
14.76
138
15.13
115
19.84
147
6.66
108
10.36
111
8.58
119
0.02
65
0.04
88
0.00
1
0.03
99
0.11
59
0.24
102
CVANet_RVCtwo views4.16
77
1.16
98
3.60
43
1.94
74
1.46
78
3.92
55
4.68
125
10.89
84
8.34
78
7.58
77
10.84
76
10.27
94
6.62
106
8.56
84
2.69
46
0.39
142
0.00
1
0.00
1
0.01
82
0.21
95
0.09
56
HSM-Net_RVCpermissivetwo views4.20
79
0.32
2
2.76
24
0.63
20
0.69
46
6.95
96
1.69
81
11.96
97
8.36
79
8.83
92
12.17
89
15.18
124
4.21
67
6.91
54
3.30
62
0.02
65
0.02
70
0.00
1
0.00
1
0.01
7
0.01
8
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
MMNettwo views4.16
77
0.77
55
5.47
84
1.40
43
1.67
86
10.78
126
0.63
52
8.64
49
8.47
80
8.80
91
8.33
55
7.51
69
6.61
105
7.53
62
6.54
98
0.00
1
0.00
1
0.00
1
0.00
1
0.03
26
0.05
43
GwcNet-DCAtwo views2.08
22
0.64
44
3.81
49
0.31
10
0.17
20
1.70
15
0.25
28
5.23
16
8.47
80
6.07
48
5.30
29
1.41
15
4.23
68
1.85
12
2.02
38
0.00
1
0.00
1
0.00
1
0.00
1
0.04
31
0.19
89
222two views2.20
26
0.61
38
5.31
80
0.31
10
0.19
22
1.69
12
0.11
12
5.37
19
8.54
82
5.80
42
5.17
23
1.44
16
5.41
91
2.00
15
1.92
34
0.00
1
0.00
1
0.00
1
0.00
1
0.04
31
0.18
86
FADNet-RVC-Resampletwo views3.79
66
1.62
122
12.06
141
1.43
45
0.66
44
5.94
83
2.41
88
10.18
70
8.58
83
6.28
56
4.22
13
5.33
44
4.80
84
7.71
64
3.19
59
0.17
119
0.21
139
0.17
145
0.12
124
0.41
126
0.29
114
ccnettwo views6.87
122
0.92
74
5.15
77
3.09
134
2.18
105
11.78
133
4.90
126
17.97
148
8.79
84
12.79
129
20.20
133
16.00
129
8.35
117
9.12
95
14.17
150
0.42
147
0.33
151
0.25
153
0.09
121
0.40
125
0.48
128
StereoDRNet-Refinedtwo views4.46
87
0.62
43
3.80
48
1.92
71
0.40
33
9.35
111
0.15
18
10.02
64
8.83
85
12.69
126
11.62
85
9.34
86
3.87
61
8.06
74
8.02
111
0.00
1
0.00
1
0.01
95
0.05
110
0.20
90
0.26
108
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
ADCReftwo views7.27
125
1.38
109
16.37
154
2.52
114
3.30
130
11.63
132
3.16
99
10.80
81
9.35
86
13.03
132
25.27
149
8.17
78
8.92
123
8.06
74
21.81
164
0.15
115
0.08
110
0.16
144
0.34
147
0.38
120
0.58
137
CFNet-pseudotwo views3.23
47
0.36
10
3.58
42
2.13
88
1.45
77
3.78
51
0.70
57
12.67
109
9.52
87
5.16
33
12.28
93
2.62
29
1.81
32
5.36
40
3.08
55
0.01
48
0.00
1
0.00
1
0.00
1
0.00
1
0.07
47
CBMV_ROBtwo views4.14
76
0.52
25
3.14
33
1.30
38
0.77
52
6.92
95
1.97
85
10.11
68
9.58
88
8.92
94
14.20
109
7.12
66
5.90
99
8.65
86
3.50
68
0.01
48
0.05
93
0.00
1
0.00
1
0.04
31
0.09
56
StereoDRNettwo views5.59
97
1.75
130
6.80
111
3.12
136
4.45
156
10.61
125
4.35
119
18.80
153
9.73
89
12.22
122
6.87
38
11.44
105
4.65
77
8.09
76
8.26
116
0.02
65
0.11
120
0.00
1
0.03
99
0.20
90
0.28
113
iResNettwo views3.68
60
0.91
71
7.94
122
2.97
132
0.34
30
4.44
66
0.48
38
7.70
40
9.74
90
7.72
78
12.74
97
4.03
35
2.87
45
8.05
73
3.37
63
0.02
65
0.01
54
0.00
1
0.00
1
0.10
53
0.09
56
Anonymous Stereotwo views6.16
106
3.15
155
23.75
166
2.97
132
2.48
116
4.39
65
13.30
168
9.21
53
9.86
91
9.56
103
8.76
58
6.79
58
1.99
33
13.50
137
13.04
146
0.01
48
0.05
93
0.00
1
0.06
113
0.22
98
0.19
89
CFNettwo views3.72
63
1.10
93
5.03
74
2.49
113
1.59
81
4.90
73
0.22
24
11.38
88
9.88
92
4.80
28
11.25
79
6.44
56
3.68
58
8.33
79
3.00
52
0.00
1
0.00
1
0.00
1
0.00
1
0.22
98
0.07
47
NCCL2two views5.88
100
1.59
120
5.44
82
1.87
66
0.92
61
9.55
114
11.55
166
12.11
99
9.94
93
9.67
104
8.85
59
22.28
155
7.41
112
8.78
88
7.17
102
0.01
48
0.00
1
0.03
115
0.00
1
0.13
72
0.23
99
aanetorigintwo views8.72
141
3.29
158
27.55
169
1.95
75
3.84
147
4.93
74
5.39
131
5.49
24
10.05
94
27.85
165
31.47
160
15.80
125
12.78
141
9.04
94
11.98
141
0.25
129
0.22
140
0.22
150
0.25
139
1.28
155
0.84
148
DRN-Testtwo views5.87
99
0.98
80
5.89
92
2.69
120
3.65
142
12.37
136
3.35
103
20.07
165
10.20
95
11.93
121
12.31
94
11.06
104
5.31
89
7.89
69
9.05
124
0.04
88
0.05
93
0.04
122
0.04
108
0.18
88
0.25
105
iResNetv2_ROBtwo views4.28
85
1.43
111
7.17
115
2.91
126
1.26
72
4.36
61
1.62
79
13.64
119
10.25
96
9.83
105
11.41
80
7.68
72
4.00
63
7.75
67
1.85
33
0.00
1
0.00
1
0.00
1
0.00
1
0.37
119
0.09
56
CC-Net-ROBtwo views7.68
128
9.21
174
7.46
119
3.82
154
3.66
143
15.68
147
5.02
127
17.95
147
10.45
97
6.44
60
14.00
106
7.63
71
4.77
80
10.75
116
3.10
57
5.99
177
2.46
175
16.28
179
1.08
166
6.30
176
1.56
161
TDLMtwo views4.11
75
1.11
94
3.54
41
1.62
57
1.04
66
3.91
54
7.41
151
10.60
76
10.67
98
6.38
57
12.59
96
5.95
53
4.77
80
8.79
89
3.04
53
0.58
152
0.00
1
0.01
95
0.00
1
0.19
89
0.12
71
SGM-Foresttwo views4.96
91
0.32
2
2.84
26
1.21
35
0.64
43
10.23
122
6.64
144
11.55
90
10.98
99
10.94
116
13.59
103
11.65
106
4.30
70
8.94
92
4.63
79
0.11
109
0.04
88
0.00
1
0.00
1
0.05
39
0.46
127
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
GMStereotwo views1.83
16
0.99
81
4.55
65
0.57
19
0.05
3
2.53
24
0.01
1
4.95
14
11.14
100
3.52
17
2.24
8
0.53
12
0.49
5
3.76
29
0.80
14
0.20
124
0.00
1
0.00
1
0.12
124
0.12
64
0.00
1
ADCMidtwo views10.24
152
3.13
153
20.70
161
2.21
94
2.39
113
11.23
130
6.19
135
14.17
125
11.19
101
23.20
159
22.25
138
17.89
137
19.54
159
18.51
158
26.21
172
0.45
149
0.42
157
1.10
169
1.29
168
1.56
160
1.18
154
ADCP+two views8.09
134
1.79
132
14.50
152
1.54
50
4.28
151
16.57
151
5.20
129
12.80
110
11.20
102
12.83
130
17.07
123
11.02
103
10.80
133
17.59
156
23.18
167
0.03
79
0.05
93
0.01
95
0.18
129
0.39
124
0.81
146
Nwc_Nettwo views6.97
124
1.25
102
6.63
108
3.82
154
3.37
131
10.83
128
1.67
80
19.56
163
11.35
103
8.36
86
23.62
142
17.19
134
11.44
136
11.21
121
8.08
114
0.80
157
0.00
1
0.00
1
0.02
89
0.13
72
0.09
56
ADCPNettwo views9.54
145
2.39
144
31.46
171
2.09
84
1.60
82
16.71
153
6.39
142
12.11
99
11.45
104
13.53
134
21.45
137
19.41
145
10.94
134
14.38
143
21.54
162
0.27
134
1.16
167
0.39
161
1.49
171
0.58
133
1.45
159
NOSS_ROBtwo views3.30
50
0.46
20
2.62
17
2.08
83
1.01
65
5.60
79
0.74
62
10.37
72
11.48
105
5.15
32
8.43
56
5.67
48
1.73
29
7.97
70
2.34
42
0.02
65
0.06
100
0.00
1
0.00
1
0.07
45
0.14
76
LALA_ROBtwo views6.58
114
1.80
133
6.25
97
1.26
36
0.94
64
10.08
119
9.02
154
16.00
139
11.51
106
12.74
127
13.02
99
24.77
157
5.25
88
10.56
114
8.02
111
0.04
88
0.05
93
0.00
1
0.02
89
0.10
53
0.25
105
FADNettwo views4.23
81
1.65
124
11.75
140
1.64
59
0.80
54
4.80
70
0.77
63
13.76
120
11.65
107
3.97
22
5.24
28
9.62
88
5.14
87
8.40
80
3.78
69
0.21
125
0.04
88
0.07
131
0.05
110
1.14
151
0.10
64
RGCtwo views6.88
123
2.23
142
6.13
96
4.05
158
4.73
163
8.94
108
2.78
94
15.19
135
11.74
108
11.13
117
19.34
132
17.86
136
10.42
131
13.02
134
8.03
113
0.73
155
0.01
54
0.24
152
0.41
151
0.31
111
0.38
119
test_xeample3two views3.34
54
0.93
76
4.33
60
0.28
9
0.12
14
3.46
42
3.98
111
6.58
33
11.84
109
6.79
64
13.56
102
3.52
32
4.23
68
3.71
28
2.05
40
0.01
48
0.01
54
0.00
1
0.00
1
0.02
14
1.40
158
csctwo views2.28
27
0.61
38
3.81
49
0.22
4
0.07
7
1.69
12
0.11
12
5.37
19
11.84
109
5.80
42
5.17
23
1.44
16
5.41
91
2.00
15
1.92
34
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.18
86
cscssctwo views2.28
27
0.61
38
3.81
49
0.22
4
0.07
7
1.69
12
0.11
12
5.37
19
11.84
109
5.80
42
5.17
23
1.44
16
5.41
91
2.00
15
1.92
34
0.00
1
0.00
1
0.00
1
0.00
1
0.02
14
0.18
86
edge stereotwo views6.76
117
1.01
82
6.76
110
2.20
93
2.45
115
6.41
88
2.45
89
14.84
130
11.98
112
15.29
139
18.31
130
22.02
154
12.56
139
10.82
117
7.49
104
0.03
79
0.06
100
0.11
137
0.03
99
0.30
108
0.14
76
Syn2CoExtwo views6.32
109
2.72
148
9.33
130
2.45
111
1.67
86
8.59
105
3.32
101
18.08
149
12.27
113
8.53
89
21.08
135
10.30
95
9.80
128
9.30
101
7.86
109
0.01
48
0.00
1
0.00
1
0.00
1
0.63
137
0.43
124
CBMVpermissivetwo views5.35
96
0.91
71
3.67
45
1.62
57
0.44
35
10.09
120
7.19
150
12.49
106
12.33
114
12.22
122
14.69
112
10.93
102
6.48
103
8.51
83
4.96
82
0.02
65
0.15
129
0.00
1
0.00
1
0.17
83
0.17
84
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
DGSMNettwo views6.47
112
2.61
146
13.64
150
3.46
146
3.74
145
8.97
109
6.01
133
14.15
124
13.01
115
7.46
75
14.10
107
9.94
90
4.79
83
10.44
112
6.22
93
1.30
166
1.70
172
1.80
173
1.52
173
2.27
166
2.35
170
PA-Nettwo views4.98
92
1.47
113
7.42
118
2.40
107
2.14
102
8.73
107
3.64
109
12.42
105
13.11
116
7.03
67
7.57
49
7.88
75
6.52
104
10.16
110
7.82
108
0.02
65
0.03
81
0.00
1
0.00
1
0.11
59
1.07
153
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
NCC-stereotwo views6.77
118
1.49
114
6.48
104
2.92
128
4.40
152
7.43
99
3.61
107
19.52
161
13.29
117
8.39
87
16.91
121
15.96
126
12.13
137
12.85
131
7.70
106
1.47
167
0.11
120
0.01
95
0.42
152
0.14
79
0.24
102
Abc-Nettwo views6.77
118
1.49
114
6.48
104
2.92
128
4.40
152
7.43
99
3.61
107
19.52
161
13.29
117
8.39
87
16.91
121
15.96
126
12.13
137
12.85
131
7.70
106
1.47
167
0.11
120
0.01
95
0.42
152
0.14
79
0.24
102
Xing Li, Yangyu Fan, Guoyun Lv, and Haoyue Ma: Area-based Correlation and Non-local Attention Network for Stereo Matching. The Visual Computer
MDST_ROBtwo views8.37
137
0.32
2
9.03
128
4.18
161
2.42
114
26.86
169
6.14
134
19.36
158
13.52
119
27.09
164
22.75
141
9.47
87
4.74
79
15.06
147
6.34
96
0.02
65
0.02
70
0.00
1
0.00
1
0.02
14
0.13
75
111two views2.78
39
0.92
74
3.81
49
0.22
4
0.07
7
2.83
32
0.70
57
6.03
27
14.00
120
3.22
14
8.93
61
5.48
46
3.30
49
3.66
27
2.03
39
0.00
1
0.00
1
0.00
1
0.00
1
0.10
53
0.29
114
WCMA_ROBtwo views9.21
143
0.87
67
7.37
117
2.54
115
2.13
101
13.59
142
5.80
132
11.64
91
14.01
121
24.43
161
32.99
165
27.09
161
18.02
153
12.51
130
9.85
132
0.81
158
0.07
107
0.01
95
0.01
82
0.16
81
0.23
99
SGM_RVCbinarytwo views10.08
149
0.60
36
3.42
38
2.30
100
0.32
29
19.41
158
6.33
140
18.95
154
14.64
122
25.14
163
24.32
146
33.34
170
18.79
157
19.86
161
12.55
145
0.25
129
0.26
148
0.22
150
0.24
137
0.34
116
0.40
121
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
SGM+DAISYtwo views15.62
166
7.26
171
19.28
159
8.94
174
10.11
178
26.25
168
10.49
163
19.36
158
14.65
123
30.64
168
33.59
166
33.00
169
22.32
168
24.96
170
16.42
154
7.90
180
6.25
181
4.51
177
3.37
177
5.86
174
7.20
177
MSMD_ROBtwo views9.28
144
1.09
91
4.65
69
1.58
53
0.39
32
16.52
150
4.41
120
13.60
117
14.87
124
22.34
153
39.89
173
25.67
159
20.71
162
12.42
129
6.98
100
0.34
139
0.03
81
0.00
1
0.00
1
0.05
39
0.09
56
NVstereo2Dtwo views4.51
88
0.82
62
6.86
113
3.28
141
3.38
132
8.16
104
3.13
98
10.51
74
15.15
125
4.90
29
6.89
39
7.87
74
4.78
82
9.88
107
3.91
72
0.01
48
0.00
1
0.00
1
0.06
113
0.02
14
0.58
137
RPtwo views6.84
121
1.29
108
5.53
85
3.92
157
5.18
164
6.32
86
3.53
105
11.73
94
15.31
126
9.54
102
22.38
139
18.25
139
14.47
148
10.11
109
7.49
104
0.91
161
0.01
54
0.12
139
0.15
127
0.33
113
0.19
89
SPS-STEREOcopylefttwo views15.04
164
6.23
168
13.21
148
11.34
177
11.65
179
23.30
164
7.15
149
24.16
171
15.65
127
31.78
170
29.19
156
31.62
168
21.32
164
24.62
169
19.50
157
7.59
179
4.19
180
3.22
176
1.48
170
6.99
178
6.54
175
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
NaN_ROBtwo views6.00
103
1.24
101
6.29
98
1.34
39
1.68
88
9.60
116
10.31
162
15.09
131
15.79
128
12.62
125
8.95
62
11.67
107
5.83
97
11.78
125
6.41
97
0.05
96
0.13
126
0.08
132
0.20
131
0.22
98
0.79
145
MeshStereopermissivetwo views11.52
157
1.52
116
4.55
65
1.89
69
1.46
78
19.87
160
5.11
128
20.66
167
15.91
129
32.67
171
34.51
168
39.34
177
21.15
163
18.74
159
12.10
142
0.11
109
0.06
100
0.01
95
0.00
1
0.45
128
0.22
95
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
DISCOtwo views6.28
108
0.57
33
5.78
89
3.43
145
1.17
69
11.22
129
3.39
104
12.14
102
16.16
130
6.52
61
11.22
78
16.96
132
6.32
102
19.51
160
10.74
136
0.00
1
0.00
1
0.00
1
0.00
1
0.35
117
0.11
67
RYNettwo views6.34
110
0.89
70
5.88
91
1.41
44
4.48
158
15.97
149
4.18
114
13.41
115
16.49
131
10.81
115
7.00
41
14.33
120
8.72
121
9.43
104
13.71
147
0.00
1
0.01
54
0.00
1
0.00
1
0.02
14
0.07
47
AnyNet_C01two views16.12
168
10.81
177
59.36
178
4.42
163
2.49
117
30.06
171
15.15
176
17.51
146
16.51
132
17.88
144
37.69
172
24.04
156
17.54
152
29.60
174
33.29
180
0.28
135
0.38
153
0.43
162
0.42
152
2.57
168
1.98
165
pmcnntwo views7.72
130
1.27
104
9.42
131
2.91
126
3.14
126
9.44
112
6.23
137
12.56
108
16.51
132
14.53
136
24.08
144
27.44
162
8.49
119
9.32
102
8.44
118
0.06
100
0.08
110
0.00
1
0.00
1
0.30
108
0.15
79
GwcNetcopylefttwo views6.42
111
1.97
138
10.92
136
2.59
118
5.58
167
11.55
131
2.21
87
14.10
123
16.52
134
10.04
108
17.19
126
10.86
101
5.61
95
9.23
98
8.84
123
0.01
48
0.06
100
0.03
115
0.08
117
0.56
130
0.40
121
iResNet_ROBtwo views4.23
81
1.02
84
4.90
71
2.18
91
0.93
63
2.92
35
0.37
33
15.10
133
16.91
135
7.89
81
10.51
74
7.03
60
3.07
47
8.16
78
3.46
67
0.01
48
0.00
1
0.00
1
0.00
1
0.10
53
0.02
14
AnyNet_C32two views10.98
156
5.58
165
22.79
164
4.16
159
5.83
168
15.64
146
14.30
170
13.18
113
17.15
136
16.44
143
20.52
134
14.68
122
13.44
145
22.46
165
30.08
177
0.17
119
0.26
148
0.36
159
0.36
148
1.23
153
0.91
149
GANetREF_RVCpermissivetwo views6.56
113
2.89
150
7.58
121
3.41
143
0.40
33
12.96
139
9.58
158
15.09
131
17.25
137
10.33
109
10.62
75
12.27
110
8.16
116
12.21
127
4.53
78
0.41
144
0.00
1
0.00
1
0.02
89
3.12
169
0.39
120
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
AANet_RVCtwo views5.01
93
1.74
129
6.38
102
1.96
76
1.29
74
2.26
19
1.69
81
10.07
65
18.53
138
7.88
80
18.15
129
8.49
82
2.70
41
10.59
115
7.04
101
0.96
162
0.15
129
0.02
107
0.00
1
0.13
72
0.12
71
FAT-Stereotwo views6.78
120
0.68
49
6.80
111
2.30
100
1.77
94
5.63
80
4.20
115
18.79
152
18.62
139
10.53
111
17.15
124
18.52
141
13.74
147
8.86
91
7.38
103
0.03
79
0.15
129
0.01
95
0.07
115
0.12
64
0.26
108
FBW_ROBtwo views8.50
140
1.03
86
7.98
123
1.93
73
1.28
73
13.10
140
6.23
137
22.50
169
18.98
140
18.82
146
14.91
114
19.06
144
10.04
129
18.41
157
9.83
130
0.62
154
0.22
140
1.82
174
0.82
162
0.99
149
1.36
157
PASMtwo views7.90
132
4.22
161
21.97
163
3.25
140
3.29
129
5.39
77
6.57
143
10.57
75
19.09
141
12.77
128
13.92
105
18.11
138
9.51
127
13.79
140
10.77
138
0.19
122
0.45
159
0.29
154
1.08
166
1.49
158
1.19
155
FCDSN-DCtwo views10.24
152
0.56
32
3.49
39
1.96
76
1.29
74
16.90
154
4.59
124
17.16
145
19.10
142
24.64
162
32.46
164
33.82
171
22.14
167
15.93
151
10.45
133
0.04
88
0.00
1
0.00
1
0.00
1
0.05
39
0.21
93
Dominik Hirner, Friedrich Fraundorfer: FCDSN-DC: An accurate but lightweight end-to-end trainable neural network for stereo estimation with depth completion.
PWC_ROBbinarytwo views8.24
136
3.13
153
12.74
145
2.43
110
4.43
155
7.51
102
1.22
72
16.63
142
19.24
143
16.08
141
28.29
152
13.99
119
10.16
130
13.63
139
14.06
149
0.42
147
0.00
1
0.05
125
0.00
1
0.59
134
0.27
111
MFN_U_SF_DS_RVCtwo views9.78
148
4.27
162
14.47
151
2.29
99
2.85
124
23.40
165
13.62
169
13.60
117
19.54
144
19.42
148
24.27
145
16.74
131
8.59
120
17.05
155
7.98
110
1.25
165
1.68
171
0.17
145
2.63
174
0.72
141
1.04
151
stereogantwo views7.69
129
0.88
68
7.08
114
3.49
147
3.93
148
18.98
157
3.23
100
16.52
141
19.58
145
9.93
107
18.92
131
20.50
151
9.04
125
14.07
141
6.14
92
0.26
131
0.04
88
0.21
149
0.03
99
0.63
137
0.33
116
FC-DCNNcopylefttwo views10.72
155
0.52
25
4.27
58
1.88
67
1.63
83
17.18
155
5.29
130
18.20
150
19.69
146
28.50
166
34.51
168
34.03
172
21.48
165
15.89
150
11.15
140
0.03
79
0.01
54
0.02
107
0.01
82
0.07
45
0.09
56
MFN_U_SF_RVCtwo views12.94
158
3.66
160
25.81
168
3.61
148
2.26
109
22.77
163
4.55
123
27.10
174
20.06
147
23.90
160
28.99
154
30.53
167
16.98
151
19.92
162
20.26
158
1.24
164
1.07
166
0.98
168
1.33
169
1.80
161
2.04
166
DeepPrunerFtwo views6.75
116
2.69
147
23.31
165
3.68
149
7.16
172
3.78
51
4.29
117
13.42
116
20.13
148
8.13
84
10.46
72
7.18
68
8.06
115
11.10
120
9.44
126
0.24
128
0.15
129
0.29
154
0.42
152
0.66
139
0.45
125
AF-Nettwo views7.78
131
1.44
112
6.68
109
3.37
142
4.50
159
8.61
106
2.69
92
17.07
144
20.17
149
9.52
101
24.02
143
20.31
150
14.59
149
11.58
124
9.84
131
0.61
153
0.00
1
0.12
139
0.00
1
0.38
120
0.12
71
CSANtwo views7.62
127
1.60
121
6.56
107
1.83
64
0.66
44
12.40
137
10.52
164
14.45
127
21.32
150
14.19
135
15.98
119
17.84
135
13.02
144
12.32
128
8.38
117
0.09
106
0.07
107
0.03
115
0.04
108
0.33
113
0.67
142
LRCNet_RVCtwo views66.23
183
65.67
183
67.63
179
71.47
183
54.32
183
32.17
174
2.66
91
63.19
183
21.35
151
81.79
183
91.26
184
83.62
183
71.62
183
69.32
183
52.57
182
95.49
184
64.45
183
95.07
184
95.19
184
71.52
183
74.31
183
FINETtwo views5.32
95
1.14
96
10.04
133
2.17
90
3.67
144
6.49
90
6.76
146
11.68
92
22.08
152
7.42
74
12.15
88
4.24
37
3.61
56
9.25
99
5.35
86
0.04
88
0.04
88
0.00
1
0.00
1
0.17
83
0.09
56
S-Stereotwo views6.63
115
0.84
64
9.67
132
3.15
137
3.48
136
6.49
90
6.22
136
12.99
112
22.84
153
9.48
100
15.51
116
12.00
108
8.43
118
8.04
72
10.70
135
0.12
112
0.17
134
0.00
1
0.38
149
0.13
72
1.92
164
LSMtwo views14.01
162
5.95
166
33.49
172
6.78
173
43.61
181
10.22
121
9.98
161
15.16
134
22.93
154
23.07
158
32.34
163
18.52
141
12.67
140
15.45
149
11.10
139
0.16
118
0.51
163
0.09
135
0.32
144
1.08
150
16.85
181
ADCLtwo views10.16
151
2.11
140
19.36
160
1.92
71
1.88
98
22.23
162
8.91
153
14.04
122
23.56
155
14.62
137
26.19
150
12.75
112
13.59
146
16.06
152
22.95
166
0.26
131
0.18
136
0.75
164
0.65
158
0.69
140
0.58
137
SAMSARAtwo views14.63
163
2.74
149
12.38
142
12.65
178
6.74
171
36.50
176
72.93
184
19.36
158
23.77
156
16.20
142
13.04
100
29.21
164
12.78
141
16.98
154
15.21
152
0.11
109
0.26
148
0.03
115
0.14
126
0.76
142
0.77
144
NVStereoNet_ROBtwo views16.04
167
6.75
169
12.90
146
6.37
172
7.42
175
12.89
138
9.74
159
22.78
170
25.12
157
30.32
167
46.19
179
34.37
173
25.38
170
21.48
164
21.38
160
5.94
176
3.10
178
6.07
178
10.09
181
4.01
171
8.54
179
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
XQCtwo views8.43
139
3.58
159
16.40
155
2.92
128
2.17
103
13.22
141
3.60
106
14.64
129
25.86
158
11.87
120
12.04
86
15.06
123
10.67
132
15.24
148
19.41
156
0.39
142
0.08
110
0.05
125
0.07
115
0.84
145
0.45
125
DispFullNettwo views17.47
174
26.01
179
33.98
173
22.58
180
20.86
180
13.84
144
1.28
74
16.50
140
26.27
159
19.97
151
17.17
125
20.52
152
18.49
155
22.86
167
10.76
137
5.13
174
2.83
177
30.72
181
7.72
179
20.86
180
11.01
180
DPSNet