Low-res two-view results

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):

bad 0.5, 1.0, 2.0, 4.0 [%]: Fraction of pixels with errors larger than the given number of disparities.
Average error [px]: The per-pixel average disparity error.
Root mean square error [px]: The per-pixel root mean square disparity error.
50%, 90%, 95%, 99% error quantile [px]: The highest disparity error within the given percentage of best pixels (for 50%, this is the median error).
Time [s]: The runtime of the method.

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	all	lakes. 1l	lakes. 1s	sand box 1l	sand box 1s	stora. room 1l	stora. room 1s	stora. room 2l	stora. room 2s	stora. room 2 1l	stora. room 2 1s	stora. room 2 2l	stora. room 2 2s	stora. room 3l	stora. room 3s	tunnel 1l	tunnel 1s	tunnel 2l	tunnel 2s	tunnel 3l	tunnel 3s

R-Stereo	2.44 3	0.32 1	1.93 1	0.94 4	0.16 4	3.67 8	0.61 16	6.37 3	3.08 1	9.14 38	17.44 65	1.80 3	0.77 3	1.76 1	0.70 1	0.00 1	0.01 25	0.00 1	0.00 1	0.01 1	0.03 9

R-Stereo Train	2.44 3	0.32 1	1.93 1	0.94 4	0.16 4	3.67 8	0.61 16	6.37 3	3.08 1	9.14 38	17.44 65	1.80 3	0.77 3	1.76 1	0.70 1	0.00 1	0.01 25	0.00 1	0.00 1	0.01 1	0.03 9

HSM-Net_RVC	4.20 29	0.32 1	2.76 6	0.63 3	0.69 20	6.95 42	1.69 37	11.96 40	8.36 30	8.83 36	12.17 39	15.18 63	4.21 29	6.91 12	3.30 20	0.02 23	0.02 35	0.00 1	0.00 1	0.01 1	0.01 2
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
NVstereo2D	4.51 34	0.82 19	6.86 58	3.28 75	3.38 78	8.16 47	3.13 45	10.51 28	15.15 64	4.90 8	6.89 8	7.87 25	4.78 37	9.88 46	3.91 30	0.01 18	0.00 1	0.00 1	0.06 64	0.02 4	0.58 77

STTStereo	3.60 15	0.93 27	6.34 49	2.71 62	2.23 58	3.68 10	0.63 19	9.42 13	6.73 12	9.87 46	6.97 9	8.84 30	3.65 20	6.85 11	3.04 15	0.00 1	0.02 35	0.01 43	0.00 1	0.02 4	0.02 3

HSM	4.00 25	0.79 18	3.16 12	1.59 20	2.17 56	6.77 40	1.11 28	12.28 46	6.35 10	6.75 20	8.11 16	13.90 55	5.37 43	8.85 37	2.71 11	0.00 1	0.00 1	0.00 1	0.00 1	0.02 4	0.02 3

RYNet	6.34 52	0.89 23	5.88 40	1.41 13	4.48 93	15.97 81	4.18 57	13.41 55	16.49 70	10.81 52	7.00 10	14.33 58	8.72 61	9.43 43	13.71 83	0.00 1	0.01 25	0.00 1	0.00 1	0.02 4	0.07 18

MDST_ROB	8.37 72	0.32 1	9.03 68	4.18 92	2.42 63	26.86 101	6.14 70	19.36 90	13.52 58	27.09 99	22.75 76	9.47 33	4.74 35	15.06 80	6.34 40	0.02 23	0.02 35	0.00 1	0.00 1	0.02 4	0.13 38

AdaStereo	3.09 8	0.58 12	3.04 9	2.84 64	0.48 15	4.08 15	1.29 32	12.16 45	7.77 23	6.03 12	9.62 24	5.79 14	1.53 7	4.56 5	1.93 7	0.00 1	0.00 1	0.00 1	0.00 1	0.02 4	0.02 3
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.
DLCB_ROB	4.51 34	0.91 24	3.78 17	2.19 45	1.07 36	6.28 35	3.09 44	9.78 16	7.72 21	10.65 50	12.97 46	13.91 56	3.71 22	8.72 34	5.30 36	0.00 1	0.00 1	0.00 1	0.00 1	0.03 10	0.10 30

RASNet	4.52 36	0.61 14	4.42 24	3.42 78	4.68 96	4.58 23	0.99 26	9.54 15	8.01 24	5.28 10	11.42 36	10.34 39	8.88 62	9.28 41	8.68 60	0.15 57	0.00 1	0.00 1	0.00 1	0.03 10	0.04 15

PMTNet	1.99 2	0.32 1	2.21 3	0.39 2	0.23 6	5.08 27	0.49 11	5.84 1	8.22 27	3.07 1	3.29 1	0.73 2	0.75 2	8.18 28	0.94 3	0.00 1	0.00 1	0.00 1	0.00 1	0.03 10	0.00 1

NLCA_NET_v2_RVC	3.84 21	1.06 34	5.23 31	2.72 63	3.27 74	4.36 18	0.61 16	10.71 31	7.56 20	8.75 34	7.89 14	9.86 36	3.90 24	7.15 13	3.44 23	0.14 55	0.02 35	0.02 55	0.03 50	0.04 13	0.03 9
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
CBMV_ROB	4.14 27	0.52 9	3.14 11	1.30 10	0.77 25	6.92 41	1.97 39	10.11 22	9.58 35	8.92 37	14.20 53	7.12 22	5.90 46	8.65 33	3.50 25	0.01 18	0.05 48	0.00 1	0.00 1	0.04 13	0.09 23

SGM-Forest	4.96 37	0.32 1	2.84 7	1.21 8	0.64 17	10.23 62	6.64 78	11.55 36	10.98 44	10.94 53	13.59 49	11.65 46	4.30 30	8.94 38	4.63 33	0.11 51	0.04 44	0.00 1	0.00 1	0.05 15	0.46 69
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
MSMD_ROB	9.28 79	1.09 38	4.65 27	1.58 19	0.39 10	16.52 82	4.41 61	13.60 57	14.87 63	22.34 89	39.89 105	25.67 93	20.71 98	12.42 64	6.98 43	0.34 75	0.03 40	0.00 1	0.00 1	0.05 15	0.09 23

CC-Net-ROB	3.84 21	1.07 35	5.23 31	2.65 59	2.96 72	4.22 16	0.69 22	10.43 27	7.72 21	8.78 35	8.29 17	9.61 34	4.02 26	7.16 14	3.65 26	0.13 54	0.03 40	0.02 55	0.03 50	0.05 15	0.03 9

DPM-Stereo	1.97 1	0.64 16	2.95 8	0.17 1	0.10 1	4.83 25	0.13 2	8.60 9	4.06 4	6.42 16	4.92 4	0.44 1	0.72 1	3.57 3	1.80 5	0.00 1	0.01 25	0.00 1	0.00 1	0.05 15	0.04 15

HITNet	2.79 6	0.77 17	4.02 19	2.03 40	0.11 3	5.58 31	0.59 14	9.24 12	5.15 8	6.42 16	7.26 11	3.66 6	2.92 16	4.07 4	3.87 29	0.00 1	0.00 1	0.00 1	0.00 1	0.06 19	0.02 3
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
NOSS_ROB	3.30 11	0.46 8	2.62 4	2.08 41	1.01 34	5.60 32	0.74 23	10.37 26	11.48 49	5.15 9	8.43 18	5.67 13	1.73 8	7.97 20	2.34 9	0.02 23	0.06 54	0.00 1	0.00 1	0.07 20	0.14 39

FC-DCNN	10.72 89	0.52 9	4.27 23	1.88 32	1.63 46	17.18 86	5.29 68	18.20 84	19.69 82	28.50 100	34.51 101	34.03 104	21.48 101	15.89 83	11.15 77	0.03 33	0.01 25	0.02 55	0.01 38	0.07 20	0.09 23

ETE_ROB	5.80 43	1.77 72	6.33 47	1.44 15	0.78 26	6.43 39	6.90 79	12.53 49	8.08 25	12.93 67	14.89 55	21.13 87	5.87 45	9.83 45	6.57 42	0.04 38	0.01 25	0.00 1	0.02 42	0.08 22	0.33 61

MLCV	3.44 13	0.88 21	5.60 36	1.39 12	0.25 7	4.36 18	0.33 6	7.25 5	7.28 16	9.17 40	12.24 42	5.09 10	2.47 12	9.15 40	3.23 19	0.00 1	0.00 1	0.00 1	0.00 1	0.10 23	0.02 3

iResNet_ROB	4.23 30	1.02 31	4.90 28	2.18 44	0.93 32	2.92 5	0.37 8	15.10 71	16.91 73	7.89 27	10.51 28	7.03 19	3.07 17	8.16 27	3.46 24	0.01 18	0.00 1	0.00 1	0.00 1	0.10 23	0.02 3

iResNet	3.68 17	0.91 24	7.94 65	2.97 70	0.34 9	4.44 22	0.48 10	7.70 6	9.74 37	7.72 24	12.74 45	4.03 7	2.87 15	8.05 22	3.37 21	0.02 23	0.01 25	0.00 1	0.00 1	0.10 23	0.09 23

LALA_ROB	6.58 54	1.80 74	6.25 45	1.26 9	0.94 33	10.08 59	9.02 85	16.00 77	11.51 50	12.74 64	13.02 47	24.77 91	5.25 41	10.56 51	8.02 52	0.04 38	0.05 48	0.00 1	0.02 42	0.10 23	0.25 54

PA-Net	4.98 38	1.47 57	7.42 62	2.40 52	2.14 55	8.73 49	3.64 54	12.42 47	13.11 55	7.03 22	7.57 12	7.88 26	6.52 49	10.16 48	7.82 50	0.02 23	0.03 40	0.00 1	0.00 1	0.11 27	1.07 92
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
XPNet_ROB	6.03 48	1.22 44	5.61 37	2.56 58	0.90 29	6.32 36	7.07 80	12.92 52	8.30 28	14.76 74	15.13 57	19.84 82	6.66 52	10.36 49	8.58 59	0.02 23	0.04 44	0.00 1	0.03 50	0.11 27	0.24 51

LE_ROB	16.73 105	1.28 51	11.61 76	3.72 84	1.65 47	16.67 84	9.17 86	14.39 64	55.91 113	63.81 113	40.86 108	35.94 106	37.73 112	14.24 76	26.87 105	0.05 43	0.10 65	0.13 79	0.22 78	0.12 29	0.15 41

GANet	6.22 50	1.07 35	4.07 21	2.27 48	0.89 28	9.19 51	9.52 88	12.02 41	8.13 26	10.72 51	29.09 91	13.86 54	7.52 56	11.00 56	4.39 31	0.36 76	0.00 1	0.02 55	0.02 42	0.12 29	0.08 21

AANet_RVC	5.01 39	1.74 70	6.38 50	1.96 39	1.29 41	2.26 3	1.69 37	10.07 18	18.53 76	7.88 26	18.15 67	8.49 29	2.70 14	10.59 52	7.04 44	0.96 98	0.15 76	0.02 55	0.00 1	0.13 31	0.12 33

DANet	6.02 47	1.23 45	8.45 67	3.86 87	3.94 85	7.64 46	1.34 34	9.51 14	7.00 13	13.39 69	15.53 58	15.99 66	7.02 53	12.14 61	12.37 79	0.19 62	0.12 72	0.02 55	0.03 50	0.13 31	0.56 76

Nwc_Net	6.97 61	1.25 47	6.63 54	3.82 86	3.37 77	10.83 65	1.67 36	19.56 95	11.35 47	8.36 29	23.62 77	17.19 71	11.44 73	11.21 58	8.08 55	0.80 93	0.00 1	0.00 1	0.02 42	0.13 31	0.09 23

NCCL2	5.88 45	1.59 64	5.44 34	1.87 31	0.92 31	9.55 55	11.55 97	12.11 42	9.94 40	9.67 44	8.85 21	22.28 89	7.41 54	8.78 35	7.17 45	0.01 18	0.00 1	0.03 61	0.00 1	0.13 31	0.23 48

Abc-Net	6.77 57	1.49 58	6.48 51	2.92 67	4.40 87	7.43 43	3.61 52	19.52 93	13.29 56	8.39 30	16.91 61	15.96 64	12.13 74	12.85 66	7.70 48	1.47 102	0.11 67	0.01 43	0.42 91	0.14 35	0.24 51

NCC-stereo	6.77 57	1.49 58	6.48 51	2.92 67	4.40 87	7.43 43	3.61 52	19.52 93	13.29 56	8.39 30	16.91 61	15.96 64	12.13 74	12.85 66	7.70 48	1.47 102	0.11 67	0.01 43	0.42 91	0.14 35	0.24 51

WCMA_ROB	9.21 78	0.87 20	7.37 61	2.54 57	2.13 54	13.59 76	5.80 69	11.64 37	14.01 60	24.43 97	32.99 98	27.09 95	18.02 89	12.51 65	9.85 71	0.81 94	0.07 58	0.01 43	0.01 38	0.16 37	0.23 48

CBMV	5.35 41	0.91 24	3.67 16	1.62 22	0.44 13	10.09 60	7.19 82	12.49 48	12.33 54	12.22 59	14.69 54	10.93 42	6.48 48	8.51 31	4.96 34	0.02 23	0.15 76	0.00 1	0.00 1	0.17 38	0.17 43
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
DRN-Test	5.87 44	0.98 29	5.89 41	2.69 61	3.65 83	12.37 70	3.35 48	20.07 97	10.20 41	11.93 58	12.31 43	11.06 44	5.31 42	7.89 19	9.05 63	0.04 38	0.05 48	0.04 66	0.04 59	0.18 39	0.25 54

DMCA	3.29 10	1.05 33	4.18 22	1.60 21	2.87 71	2.11 2	0.60 15	7.95 7	4.65 7	8.62 33	8.59 19	9.24 31	4.68 34	6.25 8	3.03 14	0.04 38	0.09 64	0.06 70	0.03 50	0.18 39	0.07 18

DMCA-RVC	3.68 17	1.58 63	6.33 47	1.79 28	2.33 60	4.04 14	0.68 20	9.07 10	4.57 6	8.48 32	9.34 23	10.91 41	3.45 19	6.67 10	3.78 27	0.02 23	0.11 67	0.01 43	0.04 59	0.18 39	0.12 33

TDLM	4.11 26	1.11 40	3.54 14	1.62 22	1.04 35	3.91 12	7.41 83	10.60 30	10.67 43	6.38 15	12.59 44	5.95 15	4.77 36	8.79 36	3.04 15	0.58 88	0.00 1	0.01 43	0.00 1	0.19 42	0.12 33

StereoDRNet-Refined	4.46 33	0.62 15	3.80 18	1.92 35	0.40 11	9.35 52	0.15 3	10.02 17	8.83 32	12.69 63	11.62 37	9.34 32	3.87 23	8.06 23	8.02 52	0.00 1	0.00 1	0.01 43	0.05 62	0.20 43	0.26 57
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
StereoDRNet	5.59 42	1.75 71	6.80 57	3.12 72	4.45 91	10.61 64	4.35 60	18.80 86	9.73 36	12.22 59	6.87 7	11.44 45	4.65 33	8.09 25	8.26 56	0.02 23	0.11 67	0.00 1	0.03 50	0.20 43	0.28 59

ccs_rob	3.63 16	1.12 41	4.42 24	2.52 55	0.91 30	5.50 30	0.21 4	10.11 22	9.11 33	6.55 19	11.28 34	8.32 28	2.55 13	7.66 15	2.01 8	0.00 1	0.00 1	0.00 1	0.00 1	0.20 43	0.08 21

PSMNet_ROB	5.02 40	1.63 67	6.03 43	1.90 34	1.83 52	9.57 56	6.35 75	15.58 76	7.23 15	6.15 13	10.48 27	12.22 48	4.16 28	8.02 21	8.71 61	0.02 23	0.01 25	0.01 43	0.10 70	0.20 43	0.12 33

ccs	3.04 7	0.39 7	3.08 10	1.78 27	0.52 16	2.04 1	0.50 12	13.09 53	13.71 59	3.54 4	5.36 6	5.50 12	2.45 11	4.81 6	2.88 12	0.09 48	0.08 60	0.12 76	0.10 70	0.20 43	0.50 72

CVANet_RVC	4.16 28	1.16 43	3.60 15	1.94 38	1.46 42	3.92 13	4.68 65	10.89 34	8.34 29	7.58 23	10.84 30	10.27 38	6.62 50	8.56 32	2.69 10	0.39 78	0.00 1	0.00 1	0.01 38	0.21 48	0.09 23

CFNet	3.72 19	1.10 39	5.03 29	2.49 54	1.59 44	4.90 26	0.22 5	11.38 35	9.88 39	4.80 7	11.25 33	6.44 17	3.68 21	8.33 29	3.00 13	0.00 1	0.00 1	0.00 1	0.00 1	0.22 49	0.07 18

Anonymous Stereo	6.16 49	3.15 92	23.75 101	2.97 70	2.48 65	4.39 21	13.30 99	9.21 11	9.86 38	9.56 43	8.76 20	6.79 18	1.99 9	13.50 71	13.04 82	0.01 18	0.05 48	0.00 1	0.06 64	0.22 49	0.19 45

NaN_ROB	6.00 46	1.24 46	6.29 46	1.34 11	1.68 49	9.60 57	10.31 93	15.09 69	15.79 67	12.62 62	8.95 22	11.67 47	5.83 44	11.78 60	6.41 41	0.05 43	0.13 73	0.08 72	0.20 77	0.22 49	0.79 85

CFNet_RVC	3.31 12	0.94 28	2.69 5	1.50 16	2.38 61	2.81 4	0.68 20	8.35 8	7.43 19	4.45 6	9.94 25	10.20 37	4.60 32	6.49 9	3.41 22	0.00 1	0.00 1	0.03 61	0.00 1	0.22 49	0.03 9

BEATNet_4x	3.24 9	1.27 48	5.89 41	1.56 18	0.10 1	5.26 28	1.07 27	10.08 19	5.50 9	6.89 21	7.73 13	4.53 9	4.13 27	5.05 7	5.27 35	0.04 38	0.05 48	0.00 1	0.00 1	0.23 53	0.23 48

SGM-ForestM	16.99 106	1.08 37	5.74 38	2.12 43	0.75 22	31.63 105	12.21 98	27.80 108	32.25 99	37.88 107	39.99 106	52.96 112	35.20 111	33.60 108	24.47 102	0.26 68	0.39 89	0.31 92	0.39 89	0.26 54	0.53 75

edge stereo	6.76 56	1.01 30	6.76 56	2.20 46	2.45 64	6.41 38	2.45 41	14.84 68	11.98 53	15.29 75	18.31 68	22.02 88	12.56 76	10.82 53	7.49 46	0.03 33	0.06 54	0.11 75	0.03 50	0.30 55	0.14 39

pmcnn	7.72 65	1.27 48	9.42 70	2.91 65	3.14 73	9.44 53	6.23 72	12.56 50	16.51 71	14.53 72	24.08 79	27.44 96	8.49 59	9.32 42	8.44 58	0.06 45	0.08 60	0.00 1	0.00 1	0.30 55	0.15 41

RGC	6.88 60	2.23 81	6.13 44	4.05 89	4.73 97	8.94 50	2.78 43	15.19 73	11.74 52	11.13 54	19.34 70	17.86 73	10.42 69	13.02 68	8.03 54	0.73 91	0.01 25	0.24 89	0.41 90	0.31 57	0.38 63

DeepPruner_ROB	3.52 14	1.14 42	4.06 20	1.12 7	1.65 47	3.65 7	0.83 25	13.96 61	4.47 5	7.80 25	10.84 30	7.05 21	2.16 10	8.14 26	3.08 17	0.07 46	0.03 40	0.00 1	0.01 38	0.32 58	0.06 17

CSAN	7.62 63	1.60 65	6.56 53	1.83 29	0.66 18	12.40 71	10.52 95	14.45 65	21.32 86	14.19 71	15.98 60	17.84 72	13.02 80	12.32 63	8.38 57	0.09 48	0.07 58	0.03 61	0.04 59	0.33 59	0.67 83

RP	6.84 59	1.29 52	5.53 35	3.92 88	5.18 98	6.32 36	3.53 50	11.73 38	15.31 65	9.54 42	22.38 75	18.25 76	14.47 83	10.11 47	7.49 46	0.91 97	0.01 25	0.12 76	0.15 74	0.33 59	0.19 45

SGM_RVC	10.08 84	0.60 13	3.42 13	2.30 50	0.32 8	19.41 89	6.33 74	18.95 87	14.64 61	25.14 98	24.32 81	33.34 103	18.79 93	19.86 93	12.55 81	0.25 67	0.26 85	0.22 88	0.24 80	0.34 61	0.40 65
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
DN-CSS_ROB	2.69 5	1.40 54	5.34 33	2.31 51	0.75 22	3.14 6	0.06 1	6.11 2	3.87 3	5.34 11	12.18 40	2.34 5	1.22 5	7.84 18	1.48 4	0.03 33	0.00 1	0.00 1	0.00 1	0.35 62	0.03 9

DISCO	6.28 51	0.57 11	5.78 39	3.43 79	1.17 37	11.22 66	3.39 49	12.14 44	16.16 69	6.52 18	11.22 32	16.96 69	6.32 47	19.51 92	10.74 73	0.00 1	0.00 1	0.00 1	0.00 1	0.35 62	0.11 32

iResNetv2_ROB	4.28 32	1.43 55	7.17 60	2.91 65	1.26 39	4.36 18	1.62 35	13.64 59	10.25 42	9.83 45	11.41 35	7.68 24	4.00 25	7.75 17	1.85 6	0.00 1	0.00 1	0.00 1	0.00 1	0.37 64	0.09 23

AF-Net	7.78 66	1.44 56	6.68 55	3.37 76	4.50 94	8.61 48	2.69 42	17.07 82	20.17 85	9.52 41	24.02 78	20.31 84	14.59 84	11.58 59	9.84 70	0.61 89	0.00 1	0.12 76	0.00 1	0.38 65	0.12 33

SHD	9.61 81	2.60 84	12.46 80	3.69 83	3.54 80	9.47 54	1.25 30	20.16 98	37.84 107	18.19 81	21.24 72	16.96 69	12.83 79	14.47 79	16.05 87	0.32 74	0.13 73	0.01 43	0.08 67	0.38 65	0.48 70

MANE	19.47 110	1.27 48	5.07 30	4.69 95	5.55 99	30.49 104	9.94 91	34.01 111	37.27 106	44.13 110	51.57 114	52.51 111	40.41 113	33.58 107	24.81 103	0.89 96	0.86 100	1.11 105	9.72 112	0.38 65	1.06 91

ADCRef	7.27 62	1.38 53	16.37 90	2.52 55	3.30 76	11.63 68	3.16 46	10.80 32	9.35 34	13.03 68	25.27 85	8.17 27	8.92 63	8.06 23	21.81 98	0.15 57	0.08 60	0.16 81	0.34 87	0.38 65	0.58 77

ADCP+	8.09 69	1.79 73	14.50 88	1.54 17	4.28 86	16.57 83	5.20 67	12.80 51	11.20 46	12.83 66	17.07 63	11.02 43	10.80 71	17.59 88	23.18 101	0.03 33	0.05 48	0.01 43	0.18 75	0.39 69	0.81 86

FADNet-RVC-Resample	3.79 20	1.62 66	12.06 78	1.43 14	0.66 18	5.94 34	2.41 40	10.18 24	8.58 31	6.28 14	4.22 3	5.33 11	4.80 38	7.71 16	3.19 18	0.17 60	0.21 81	0.17 82	0.12 72	0.41 70	0.29 60

PVD	15.44 98	2.93 88	14.67 89	4.21 93	3.39 79	17.43 87	4.16 56	27.84 109	48.84 112	31.02 103	43.54 109	29.76 99	30.81 109	25.97 103	21.40 95	0.23 65	0.41 90	0.04 66	0.33 86	0.41 70	1.33 95

MeshStereo	11.52 91	1.52 60	4.55 26	1.89 33	1.46 42	19.87 91	5.11 66	20.66 99	15.91 68	32.67 105	34.51 101	39.34 109	21.15 99	18.74 91	12.10 78	0.11 51	0.06 54	0.01 43	0.00 1	0.45 72	0.22 47
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
RTSC	9.15 77	3.00 89	13.57 86	3.72 84	1.76 51	11.82 69	0.46 9	16.95 81	36.83 104	15.80 76	15.53 58	12.91 51	7.46 55	20.01 95	21.76 97	0.31 73	0.13 73	0.01 43	0.08 67	0.57 73	0.41 66

ADCPNet	9.54 80	2.39 83	31.46 105	2.09 42	1.60 45	16.71 85	6.39 76	12.11 42	11.45 48	13.53 70	21.45 73	19.41 80	10.94 72	14.38 77	21.54 96	0.27 71	1.16 102	0.39 97	1.49 106	0.58 74	1.45 97

PWC_ROB	8.24 71	3.13 90	12.74 82	2.43 53	4.43 90	7.51 45	1.22 29	16.63 80	19.24 79	16.08 77	28.29 88	13.99 57	10.16 68	13.63 73	14.06 84	0.42 83	0.00 1	0.05 68	0.00 1	0.59 75	0.27 58

FADNet-RVC	3.98 24	1.84 75	12.48 81	1.69 25	0.44 13	4.33 17	1.31 33	11.84 39	7.15 14	3.53 3	3.50 2	10.63 40	4.43 31	9.12 39	6.25 38	0.03 33	0.10 65	0.00 1	0.03 50	0.60 76	0.25 54

SANet	10.64 88	1.86 76	10.91 72	1.76 26	0.71 21	14.62 79	9.23 87	19.18 88	37.14 105	19.22 83	27.96 87	25.86 94	19.11 94	13.02 68	10.63 72	0.08 47	0.06 54	0.03 61	0.02 42	0.62 77	0.81 86

stereogan	7.69 64	0.88 21	7.08 59	3.49 80	3.93 84	18.98 88	3.23 47	16.52 79	19.58 81	9.93 47	18.92 69	20.50 85	9.04 64	14.07 75	6.14 37	0.26 68	0.04 44	0.21 87	0.03 50	0.63 78	0.33 61

DeepPrunerF	6.75 55	2.69 85	23.31 100	3.68 82	7.16 104	3.78 11	4.29 58	13.42 56	20.13 84	8.13 28	10.46 26	7.18 23	8.06 57	11.10 57	9.44 65	0.24 66	0.15 76	0.29 90	0.42 91	0.66 79	0.45 67

ADCL	10.16 86	2.11 79	19.36 95	1.92 35	1.88 53	22.23 93	8.91 84	14.04 62	23.56 88	14.62 73	26.19 86	12.75 50	13.59 82	16.06 84	22.95 100	0.26 68	0.18 79	0.75 100	0.65 96	0.69 80	0.58 77

MFN_U_SF_DS_RVC	9.78 83	4.27 98	14.47 87	2.29 49	2.85 70	23.40 96	13.62 100	13.60 57	19.54 80	19.42 84	24.27 80	16.74 68	8.59 60	17.05 87	7.98 51	1.25 101	1.68 106	0.17 82	2.63 108	0.72 81	1.04 90

SAMSARA	14.63 96	2.74 86	12.38 79	12.65 110	6.74 103	36.50 107	72.93 115	19.36 90	23.77 89	16.20 78	13.04 48	29.21 97	12.78 78	16.98 86	15.21 86	0.11 51	0.26 85	0.03 61	0.14 73	0.76 82	0.77 84

G-Net	8.41 73	1.54 61	10.97 73	5.73 100	3.60 81	26.19 98	4.41 61	10.10 20	7.42 17	19.71 85	24.99 83	14.38 59	15.83 85	10.99 54	9.53 66	0.50 86	0.46 95	0.19 85	0.25 81	0.80 83	0.66 81

STTStereo_v2	8.41 73	1.54 61	10.97 73	5.73 100	3.60 81	26.19 98	4.41 61	10.10 20	7.42 17	19.71 85	24.99 83	14.38 59	15.83 85	10.99 54	9.53 66	0.50 86	0.46 95	0.19 85	0.25 81	0.80 83	0.66 81

XQC	8.43 75	3.58 95	16.40 91	2.92 67	2.17 56	13.22 75	3.60 51	14.64 67	25.86 91	11.87 57	12.04 38	15.06 62	10.67 70	15.24 81	19.41 90	0.39 78	0.08 60	0.05 68	0.07 66	0.84 85	0.45 67

FADNet_RVC	3.91 23	1.67 69	12.95 84	0.96 6	0.75 22	5.71 33	0.54 13	10.83 33	6.60 11	3.46 2	8.09 15	4.10 8	3.40 18	9.43 43	6.33 39	0.36 76	0.44 93	0.17 82	0.46 95	0.91 86	0.95 89

RTSA	18.87 108	9.32 108	86.48 113	4.95 97	6.10 101	42.08 110	14.70 103	15.49 74	41.06 109	22.65 91	32.32 95	13.77 52	19.54 95	37.98 109	28.96 106	0.41 80	0.23 83	0.00 1	0.02 42	0.91 86	0.50 72

RTS	18.87 108	9.32 108	86.48 113	4.95 97	6.10 101	42.08 110	14.70 103	15.49 74	41.06 109	22.65 91	32.32 95	13.77 52	19.54 95	37.98 109	28.96 106	0.41 80	0.23 83	0.00 1	0.02 42	0.91 86	0.50 72

FBW_ROB	8.50 76	1.03 32	7.98 66	1.93 37	1.28 40	13.10 74	6.23 72	22.50 101	18.98 77	18.82 82	14.91 56	19.06 79	10.04 67	18.41 89	9.83 69	0.62 90	0.22 82	1.82 108	0.82 100	0.99 89	1.36 96

LSM	14.01 95	5.95 101	33.49 106	6.78 105	43.61 113	10.22 61	9.98 92	15.16 72	22.93 87	23.07 94	32.34 97	18.52 77	12.67 77	15.45 82	11.10 76	0.16 59	0.51 98	0.09 74	0.32 85	1.08 90	16.85 113

FADNet	4.23 30	1.65 68	11.75 77	1.64 24	0.80 27	4.80 24	0.77 24	13.76 60	11.65 51	3.97 5	5.24 5	9.62 35	5.14 40	8.40 30	3.78 27	0.21 64	0.04 44	0.07 71	0.05 62	1.14 91	0.10 30

DPSNet	10.14 85	1.88 77	16.82 92	1.85 30	1.73 50	24.84 97	17.20 109	19.92 96	27.41 93	12.23 61	13.62 50	16.52 67	18.35 90	14.42 78	12.50 80	0.78 92	0.54 99	0.08 72	0.25 81	1.18 92	0.59 80

AnyNet_C32	10.98 90	5.58 100	22.79 99	4.16 90	5.83 100	15.64 80	14.30 101	13.18 54	17.15 74	16.44 79	20.52 71	14.68 61	13.44 81	22.46 97	30.08 109	0.17 60	0.26 85	0.36 95	0.36 88	1.23 93	0.91 88

ADCS	13.02 93	4.93 99	28.38 104	3.17 73	2.67 68	13.61 77	10.83 96	18.70 85	33.46 100	22.59 90	24.78 82	19.59 81	18.51 92	23.40 100	32.16 111	0.10 50	0.19 80	0.37 96	0.18 75	1.26 94	1.46 98

PWCDC_ROB	7.92 68	3.17 94	7.48 63	5.73 100	4.40 87	10.45 63	0.35 7	14.52 66	28.19 94	10.36 49	31.27 94	7.04 20	9.14 65	13.22 70	8.78 62	2.74 107	0.02 35	0.00 1	0.00 1	1.31 95	0.17 43

ELAS_RVC	16.54 103	2.26 82	10.09 71	5.50 99	4.46 92	28.28 102	16.72 108	25.55 105	33.54 101	40.19 108	40.30 107	36.68 107	30.03 107	29.40 105	20.61 93	0.98 99	1.21 103	0.86 102	0.70 98	1.39 96	2.16 102
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
ELAS	16.72 104	2.14 80	9.23 69	4.92 96	4.53 95	32.66 106	15.11 106	27.40 107	28.68 96	40.27 109	44.90 110	38.33 108	30.50 108	26.44 104	21.94 99	0.88 95	1.23 104	0.67 99	0.89 101	1.49 97	2.18 104
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
PASM	7.90 67	4.22 97	21.97 98	3.25 74	3.29 75	5.39 29	6.57 77	10.57 29	19.09 78	12.77 65	13.92 51	18.11 75	9.51 66	13.79 74	10.77 75	0.19 62	0.45 94	0.29 90	1.08 102	1.49 97	1.19 94

ADCMid	10.24 87	3.13 90	20.70 96	2.21 47	2.39 62	11.23 67	6.19 71	14.17 63	11.19 45	23.20 95	22.25 74	17.89 74	19.54 95	18.51 90	26.21 104	0.45 84	0.42 92	1.10 104	1.29 103	1.56 99	1.18 93

MFN_U_SF_RVC	12.94 92	3.66 96	25.81 103	3.61 81	2.26 59	22.77 94	4.55 64	27.10 106	20.06 83	23.90 96	28.99 90	30.53 100	16.98 87	19.92 94	20.26 92	1.24 100	1.07 101	0.98 103	1.33 104	1.80 100	2.04 101

PDISCO_ROB	9.62 82	1.99 78	11.51 75	9.88 108	9.61 109	21.48 92	3.83 55	19.33 89	28.49 95	11.27 55	14.17 52	19.92 83	5.02 39	16.35 85	9.18 64	5.28 109	0.41 90	0.14 80	0.09 69	2.05 101	2.36 105

BEATNet-Init1	23.31 111	9.03 107	41.67 108	4.17 91	2.53 67	45.68 112	19.47 110	33.43 110	38.45 108	47.59 112	49.10 112	59.31 113	41.80 114	38.35 111	29.21 108	0.47 85	0.50 97	0.81 101	0.66 97	2.10 102	1.86 99

AnyNet_C01	16.12 101	10.81 110	59.36 111	4.42 94	2.49 66	30.06 103	15.15 107	17.51 83	16.51 71	17.88 80	37.69 104	24.04 90	17.54 88	29.60 106	33.29 112	0.28 72	0.38 88	0.43 98	0.42 91	2.57 103	1.98 100

GANetREF_RVC	6.56 53	2.89 87	7.58 64	3.41 77	0.40 11	12.96 73	9.58 89	15.09 69	17.25 75	10.33 48	10.62 29	12.27 49	8.16 58	12.21 62	4.53 32	0.41 80	0.00 1	0.00 1	0.02 42	3.12 104	0.39 64
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
NVStereoNet_ROB	16.04 100	6.75 103	12.90 83	6.37 104	7.42 107	12.89 72	9.74 90	22.78 102	25.12 90	30.32 101	46.19 111	34.37 105	25.38 104	21.48 96	21.38 94	5.94 110	3.10 110	6.07 111	10.09 113	4.01 105	8.54 111
Nikolai Smolyanskiy, Alexey Kamenev, Stan Birchfield: On the Importance of Stereo for Accurate Depth Estimation: An Efficient Semi-Supervised Deep Neural Network Approach. Arxiv
MADNet+	27.07 112	33.84 112	90.97 115	20.14 111	7.47 108	48.43 113	47.10 112	35.43 112	36.46 103	20.11 88	30.05 93	25.29 92	35.08 110	45.50 113	50.28 113	2.13 104	2.00 107	1.19 107	0.76 99	4.71 106	4.43 106

MFMNet_re	13.29 94	8.60 106	18.29 93	9.75 107	7.25 106	19.65 90	14.84 105	20.71 100	30.72 98	23.03 93	28.77 89	18.85 78	26.09 106	13.55 72	9.82 68	2.44 105	1.35 105	0.34 93	0.23 79	4.78 107	6.69 108

SGM+DAISY	15.62 99	7.26 105	19.28 94	8.94 106	10.11 110	26.25 100	10.49 94	19.36 90	14.65 62	30.64 102	33.59 99	33.00 102	22.32 102	24.96 102	16.42 88	7.90 112	6.25 113	4.51 110	3.37 109	5.86 108	7.20 109

MSC_U_SF_DS_RVC	16.41 102	6.93 104	21.83 97	5.94 103	2.81 69	38.71 108	14.59 102	24.55 104	34.87 102	33.66 106	34.35 100	29.24 98	24.20 103	22.59 98	17.95 89	2.52 106	2.81 108	1.17 106	1.51 107	5.89 109	2.16 102

SPS-STEREO	15.04 97	6.23 102	13.21 85	11.34 109	11.65 111	23.30 95	7.15 81	24.16 103	15.65 66	31.78 104	29.19 92	31.62 101	21.32 100	24.62 101	19.50 91	7.59 111	4.19 112	3.22 109	1.48 105	6.99 110	6.54 107
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
SuperB	8.10 70	3.15 92	24.67 102	2.65 59	1.23 38	9.88 58	4.29 58	10.18 24	30.07 97	11.53 56	12.18 40	6.12 16	6.65 51	10.50 50	14.47 85	0.14 55	0.11 67	0.35 94	0.25 81	13.06 111	0.48 70

DispFullNet	17.47 107	26.01 111	33.98 107	22.58 112	20.86 112	13.84 78	1.28 31	16.50 78	26.27 92	19.97 87	17.17 64	20.52 86	18.49 91	22.86 99	10.76 74	5.13 108	2.83 109	30.72 113	7.72 111	20.86 112	11.01 112

PWCK	30.53 113	44.32 113	47.25 110	29.76 113	7.23 105	40.78 109	27.10 111	44.73 113	44.32 111	47.31 111	36.37 103	47.16 110	26.05 105	41.26 112	31.87 110	21.83 113	4.03 111	29.50 112	4.67 110	27.17 113	7.80 110

DPSimNet_ROB	53.45 114	64.73 114	44.39 109	53.97 114	45.39 114	53.66 114	54.83 113	55.15 114	57.87 114	64.16 114	50.83 113	63.40 114	53.34 115	46.45 114	65.81 114	63.13 114	26.54 114	57.94 114	51.11 114	45.52 114	50.69 114

MADNet++	82.84 115	82.38 115	73.57 112	87.72 115	82.97 115	93.14 115	69.15 114	86.42 115	82.50 115	93.46 115	86.70 115	86.28 115	80.92 116	88.34 115	88.84 115	86.83 115	84.17 115	72.64 115	68.92 115	80.47 115	81.42 115

MEDIAN_ROB	98.41 116	99.70 116	99.30 117	97.09 116	97.02 116	96.89 116	95.77 117	97.66 116	97.28 116	98.79 118	98.94 116	99.18 116	98.14 117	96.89 116	96.88 116	99.96 118	99.16 116	100.00 116	99.99 116	99.69 116	99.88 116

DPSM	99.95 120	100.00 118	100.00 120	99.76 117	100.00 117	100.00 118	100.00 118	100.00 119	100.00 117	100.00 119	100.00 117	100.00 119	100.00 118	100.00 119	100.00 120	99.21 116	100.00 117	100.00 116	100.00 117	99.99 117	99.95 117

DPSM_ROB	99.95 120	100.00 118	100.00 120	99.76 117	100.00 117	100.00 118	100.00 118	100.00 119	100.00 117	100.00 119	100.00 117	100.00 119	100.00 118	100.00 119	100.00 120	99.21 116	100.00 117	100.00 116	100.00 117	99.99 117	99.95 117

DGTPSM_ROB	99.90 118	100.00 118	99.99 118	99.99 119	100.00 117	100.00 118	100.00 118	99.97 117	100.00 117	98.35 116	100.00 117	99.84 117	100.00 118	99.98 117	99.99 117	99.99 119	100.00 117	100.00 116	100.00 117	100.00 119	100.00 120

AVERAGE_ROB	99.62 117	99.95 117	98.81 116	100.00 121	100.00 117	98.08 117	95.47 116	100.00 119	100.00 117	100.00 119	100.00 117	100.00 119	100.00 118	100.00 119	99.99 117	100.00 120	100.00 117	100.00 116	100.00 117	100.00 119	100.00 120

DPSMNet_ROB	99.91 119	100.00 118	99.99 118	99.99 119	100.00 117	100.00 118	100.00 118	99.98 118	100.00 117	98.35 116	100.00 117	99.84 117	100.00 118	99.98 117	99.99 117	100.00 120	100.00 117	100.00 116	100.00 117	100.00 119	100.00 120

LSM0	100.00 122	100.00 118	100.00 120	100.00 121	100.00 117	100.00 118	100.00 118	100.00 119	100.00 117	100.00 119	100.00 117	100.00 119	100.00 118	100.00 119	100.00 120	100.00 120	100.00 117	100.00 116	100.00 117	100.00 119	99.99 119

MSMDNet													1.26 6

Low-res two-view benchmark