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

MMNet	4.16 43	0.77 33	5.47 53	1.40 19	1.67 59	10.78 88	0.63 27	8.64 19	8.47 50	8.80 51	8.33 28	7.51 33	6.61 67	7.53 31	6.54 59	0.00 1	0.00 1	0.00 1	0.00 1	0.03 14	0.05 30

R-Stereo Train	2.44 6	0.32 1	1.93 2	0.94 7	0.16 5	3.67 18	0.61 24	6.37 6	3.08 1	9.14 54	17.44 87	1.80 4	0.77 3	1.76 3	0.70 2	0.00 1	0.01 39	0.00 1	0.00 1	0.01 2	0.03 17

GANet-RSSM	3.27 16	0.60 23	3.38 21	2.94 94	2.00 69	2.61 6	0.48 15	10.93 55	6.11 19	5.47 17	10.20 40	6.57 23	5.44 57	6.25 18	2.23 11	0.00 1	0.00 1	0.08 95	0.00 1	0.17 56	0.03 17

PSMNet-RSSM	3.33 20	0.77 33	3.22 20	2.12 57	1.90 68	2.33 4	0.71 33	11.17 56	5.83 16	5.90 18	11.43 52	7.04 27	4.83 51	5.96 16	3.17 22	0.00 1	0.00 1	0.02 71	0.00 1	0.08 31	0.02 7

GwcNet-RSSM	3.77 31	1.05 55	5.06 46	2.21 62	1.87 66	2.36 5	1.52 49	8.98 20	7.96 43	5.91 19	14.59 72	7.94 38	3.50 25	6.43 20	5.70 51	0.00 1	0.00 1	0.04 86	0.00 1	0.20 61	0.08 36

CFNet-ft	3.31 18	0.94 49	2.69 7	1.50 23	2.38 76	2.81 8	0.68 30	8.35 12	7.43 35	4.45 7	9.94 37	10.20 53	4.60 42	6.50 22	3.41 28	0.00 1	0.00 1	0.03 79	0.00 1	0.22 70	0.03 17

ccs_rob	3.63 25	1.12 63	4.42 36	2.52 76	0.91 37	5.50 43	0.21 6	10.11 37	9.11 53	6.55 29	11.28 49	8.32 40	2.55 16	7.66 33	2.01 10	0.00 1	0.00 1	0.00 1	0.00 1	0.20 61	0.08 36

HITNet	2.79 10	0.77 33	4.02 31	2.03 52	0.11 3	5.58 44	0.59 23	9.24 24	5.15 9	6.42 26	7.26 19	3.66 7	2.92 20	4.07 10	3.87 36	0.00 1	0.00 1	0.00 1	0.00 1	0.06 27	0.02 7
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
AdaStereo	3.09 13	0.58 22	3.04 15	2.84 87	0.48 18	4.08 25	1.29 46	12.16 67	7.77 41	6.03 20	9.62 36	5.79 19	1.53 8	4.56 11	1.93 9	0.00 1	0.00 1	0.00 1	0.00 1	0.02 7	0.02 7
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.
cf-r	3.42 21	0.81 37	3.97 30	2.41 73	2.78 86	2.85 11	0.64 29	8.44 15	6.32 20	6.25 23	11.57 53	8.97 46	3.54 27	5.93 15	3.83 35	0.00 1	0.00 1	0.00 1	0.00 1	0.09 35	0.07 33

DMCA	2.82 11	0.64 28	4.45 38	1.61 29	0.89 34	3.86 22	0.57 22	9.18 22	5.10 8	6.24 22	6.49 13	7.17 31	2.13 12	3.31 7	4.73 43	0.00 1	0.01 39	0.00 1	0.00 1	0.04 18	0.02 7

DISCO	6.28 70	0.57 21	5.78 58	3.43 106	1.17 45	11.22 91	3.39 68	12.14 66	16.16 90	6.52 28	11.22 47	16.96 91	6.32 64	19.51 119	10.74 97	0.00 1	0.00 1	0.00 1	0.00 1	0.35 85	0.11 50

iResNetv2_ROB	4.28 51	1.43 78	7.17 81	2.91 89	1.26 47	4.36 29	1.62 50	13.64 83	10.25 63	9.83 64	11.41 50	7.68 34	4.00 34	7.75 36	1.85 8	0.00 1	0.00 1	0.00 1	0.00 1	0.37 87	0.09 40

MLCV	3.44 22	0.88 42	5.60 55	1.39 18	0.25 8	4.36 29	0.33 9	7.25 9	7.28 32	9.17 56	12.24 59	5.09 12	2.47 15	9.15 61	3.23 25	0.00 1	0.00 1	0.00 1	0.00 1	0.10 36	0.02 7

DPM-Stereo	1.97 3	0.64 28	2.95 13	0.17 2	0.10 1	4.83 36	0.13 3	8.60 17	4.06 4	6.42 26	4.92 6	0.44 3	0.72 2	3.57 8	1.80 7	0.00 1	0.01 39	0.00 1	0.00 1	0.05 22	0.04 27

HSM	4.00 40	0.79 36	3.16 19	1.59 28	2.17 72	6.77 58	1.11 42	12.28 68	6.35 23	6.75 30	8.11 26	13.90 76	5.37 56	8.85 56	2.71 16	0.00 1	0.00 1	0.00 1	0.00 1	0.02 7	0.02 7

StereoDRNet-Refined	4.46 52	0.62 27	3.80 29	1.92 44	0.40 14	9.35 73	0.15 5	10.02 32	8.83 52	12.69 85	11.62 54	9.34 47	3.87 32	8.06 43	8.02 71	0.00 1	0.00 1	0.01 59	0.05 81	0.20 61	0.26 77
Rohan Chabra, Julian Straub, Chris Sweeney, Richard Newcombe, Henry Fuchs: StereoDRNet. CVPR
CFNet_RVC	3.31 18	0.94 49	2.69 7	1.50 23	2.38 76	2.81 8	0.68 30	8.35 12	7.43 35	4.45 7	9.94 37	10.20 53	4.60 42	6.49 21	3.41 28	0.00 1	0.00 1	0.03 79	0.00 1	0.22 70	0.03 17

HGLStereo	4.25 49	0.52 15	4.65 42	3.09 97	5.18 123	6.51 56	0.73 35	10.37 42	6.74 26	7.32 36	12.44 61	7.00 25	7.21 72	7.46 30	5.51 50	0.00 1	0.02 51	0.00 1	0.00 1	0.06 27	0.10 47

RYNet	6.34 71	0.89 44	5.88 59	1.41 20	4.48 117	15.97 107	4.18 77	13.41 79	16.49 91	10.81 74	7.00 18	14.33 79	8.72 81	9.43 65	13.71 108	0.00 1	0.01 39	0.00 1	0.00 1	0.02 7	0.07 33

UNet	3.97 38	0.61 25	6.50 72	1.68 34	3.22 91	9.82 79	0.31 8	9.50 26	5.32 10	7.20 34	6.35 12	10.59 57	5.91 63	6.93 25	5.49 49	0.00 1	0.00 1	0.00 1	0.00 1	0.02 7	0.01 2

RAFT-Stereo	2.44 6	0.32 1	1.93 2	0.94 7	0.16 5	3.67 18	0.61 24	6.37 6	3.08 1	9.14 54	17.44 87	1.80 4	0.77 3	1.76 3	0.70 2	0.00 1	0.01 39	0.00 1	0.00 1	0.01 2	0.03 17
Lahav Lipson, Zachary Teed, and Jia Deng: RAFT-Stereo: Multilevel Recurrent Field Transforms for Stereo Matching. 3DV
delet	4.27 50	0.72 32	4.50 39	1.07 11	3.75 106	10.79 89	4.04 75	9.95 31	7.89 42	8.09 43	8.89 32	8.46 41	3.81 31	7.56 32	5.80 52	0.00 1	0.00 1	0.00 1	0.00 1	0.11 40	0.02 7

acv_ft	3.69 28	0.53 18	4.58 41	2.31 69	4.71 121	7.44 65	0.86 39	8.54 16	5.70 14	9.29 58	9.26 34	5.77 17	4.15 37	2.24 5	8.13 75	0.00 1	0.00 1	0.00 1	0.00 1	0.29 77	0.01 2

STTStereo	3.60 24	0.93 48	6.34 67	2.71 84	2.23 74	3.68 20	0.63 27	9.42 25	6.73 25	9.87 65	6.97 17	8.84 45	3.65 28	6.85 23	3.04 19	0.00 1	0.02 51	0.01 59	0.00 1	0.02 7	0.02 7

CREStereo	0.98 1	0.42 10	1.86 1	0.07 1	0.22 7	1.65 1	0.02 1	4.07 2	6.33 21	0.72 1	1.54 2	0.43 2	0.64 1	1.22 1	0.44 1	0.00 1	0.00 1	0.00 1	0.00 1	0.01 2	0.00 1

DLCB_ROB	4.51 53	0.91 45	3.78 28	2.19 60	1.07 44	6.28 50	3.09 62	9.78 29	7.72 39	10.65 71	12.97 64	13.91 77	3.71 30	8.72 53	5.30 48	0.00 1	0.00 1	0.00 1	0.00 1	0.03 14	0.10 47

CFNet	3.72 30	1.10 61	5.03 45	2.49 75	1.59 54	4.90 38	0.22 7	11.38 57	9.88 59	4.80 10	11.25 48	6.44 22	3.68 29	8.33 48	3.00 18	0.00 1	0.00 1	0.00 1	0.00 1	0.22 70	0.07 33

ACVNet	2.58 8	0.53 18	2.69 7	0.84 5	0.59 20	2.61 6	0.55 19	8.35 12	5.70 14	4.45 7	9.26 34	5.77 17	3.50 25	2.24 5	4.41 39	0.00 1	0.00 1	0.00 1	0.00 1	0.17 56	0.01 2

Gwc-CoAtRS	2.02 4	0.33 6	2.61 5	1.04 10	0.30 10	3.56 16	0.40 13	5.89 4	5.94 17	5.08 13	5.01 7	5.55 15	1.72 9	1.60 2	0.91 5	0.01 30	0.09 86	0.00 1	0.25 103	0.01 2	0.02 7

iResNet_ROB	4.23 47	1.02 53	4.90 44	2.18 59	0.93 40	2.92 12	0.37 12	15.10 96	16.91 95	7.89 41	10.51 43	7.03 26	3.07 21	8.16 47	3.46 31	0.01 30	0.00 1	0.00 1	0.00 1	0.10 36	0.02 7

NVstereo2D	4.51 53	0.82 38	6.86 79	3.28 102	3.38 96	8.16 68	3.13 63	10.51 45	15.15 85	4.90 11	6.89 16	7.87 35	4.78 49	9.88 68	3.91 37	0.01 30	0.00 1	0.00 1	0.06 83	0.02 7	0.58 100

NCCL2	5.88 64	1.59 87	5.44 51	1.87 40	0.92 38	9.55 76	11.55 124	12.11 64	9.94 60	9.67 63	8.85 31	22.28 114	7.41 73	8.78 54	7.17 63	0.01 30	0.00 1	0.03 79	0.00 1	0.13 47	0.23 68

GwcNet	6.42 72	1.97 101	10.92 95	2.59 80	5.58 126	11.55 93	2.21 56	14.10 87	16.52 94	10.04 67	17.19 86	10.86 60	5.61 58	9.23 62	8.84 83	0.01 30	0.06 73	0.03 79	0.08 87	0.56 96	0.40 86

UPFNet	3.82 33	0.47 12	4.34 35	2.21 62	3.38 96	7.26 62	2.69 59	9.89 30	5.96 18	7.98 42	7.53 21	7.89 37	3.15 22	7.72 35	5.88 53	0.01 30	0.02 51	0.00 1	0.00 1	0.04 18	0.06 31

CBMV_ROB	4.14 42	0.52 15	3.14 18	1.30 15	0.77 30	6.92 59	1.97 55	10.11 37	9.58 55	8.92 53	14.20 71	7.12 30	5.90 62	8.65 52	3.50 32	0.01 30	0.05 67	0.00 1	0.00 1	0.04 18	0.09 40

Anonymous Stereo	6.16 68	3.15 116	23.75 125	2.97 95	2.48 81	4.39 32	13.30 126	9.21 23	9.86 58	9.56 62	8.76 30	6.79 24	1.99 11	13.50 96	13.04 107	0.01 30	0.05 67	0.00 1	0.06 83	0.22 70	0.19 63

PA-Net	4.98 57	1.47 80	7.42 83	2.40 72	2.14 71	8.73 70	3.64 73	12.42 69	13.11 76	7.03 32	7.57 22	7.88 36	6.52 66	10.16 71	7.82 69	0.02 38	0.03 58	0.00 1	0.00 1	0.11 40	1.07 116
Zhibo Rao, Mingyi He, Yuchao Dai, Zhelun Shen: Patch Attention Network with Generative Adversarial Model for Semi-Supervised Binocular Disparity Prediction.
HSM-Net_RVC	4.20 45	0.32 1	2.76 10	0.63 4	0.69 24	6.95 60	1.69 52	11.96 62	8.36 49	8.83 52	12.17 56	15.18 84	4.21 39	6.91 24	3.30 26	0.02 38	0.02 51	0.00 1	0.00 1	0.01 2	0.01 2
Gengshan Yang, Joshua Manela, Michael Happold, and Deva Ramanan: Hierarchical Deep Stereo Matching on High-resolution Images. CVPR 2019
XPNet_ROB	6.03 67	1.22 66	5.61 56	2.56 79	0.90 36	6.32 51	7.07 106	12.92 74	8.30 47	14.76 96	15.13 76	19.84 106	6.66 70	10.36 72	8.58 79	0.02 38	0.04 63	0.00 1	0.03 70	0.11 40	0.24 71

PSMNet_ROB	5.02 59	1.63 90	6.03 62	1.90 43	1.83 65	9.57 77	6.35 101	15.58 101	7.23 31	6.15 21	10.48 42	12.22 68	4.16 38	8.02 40	8.71 81	0.02 38	0.01 39	0.01 59	0.10 91	0.20 61	0.12 51

NOSS_ROB	3.30 17	0.46 11	2.62 6	2.08 54	1.01 42	5.60 45	0.74 36	10.37 42	11.48 70	5.15 14	8.43 29	5.67 16	1.73 10	7.97 39	2.34 12	0.02 38	0.06 73	0.00 1	0.00 1	0.07 29	0.14 56

DSFCA	4.21 46	0.50 14	5.45 52	1.34 16	1.68 60	7.04 61	4.51 85	10.73 50	7.00 27	10.78 73	6.80 14	8.48 43	4.63 44	9.91 69	5.12 46	0.02 38	0.06 73	0.02 71	0.03 70	0.08 31	0.03 17

StereoDRNet	5.59 61	1.75 94	6.80 77	3.12 98	4.45 115	10.61 87	4.35 81	18.80 113	9.73 56	12.22 81	6.87 15	11.44 64	4.65 45	8.09 45	8.26 76	0.02 38	0.11 89	0.00 1	0.03 70	0.20 61	0.28 80

iResNet	3.68 27	0.91 45	7.94 86	2.97 95	0.34 12	4.44 33	0.48 15	7.70 10	9.74 57	7.72 38	12.74 63	4.03 8	2.87 19	8.05 42	3.37 27	0.02 38	0.01 39	0.00 1	0.00 1	0.10 36	0.09 40

MDST_ROB	8.37 95	0.32 1	9.03 89	4.18 119	2.42 79	26.86 128	6.14 94	19.36 117	13.52 79	27.09 123	22.75 99	9.47 48	4.74 47	15.06 106	6.34 57	0.02 38	0.02 51	0.00 1	0.00 1	0.02 7	0.13 55

FENet	2.19 5	0.34 7	3.04 15	2.05 53	0.26 9	2.82 10	0.13 3	7.02 8	5.55 13	4.97 12	1.94 3	4.91 11	3.39 23	4.71 12	2.50 13	0.02 38	0.03 58	0.00 1	0.00 1	0.00 1	0.21 65

CBMV	5.35 60	0.91 45	3.67 27	1.62 30	0.44 16	10.09 82	7.19 108	12.49 70	12.33 75	12.22 81	14.69 73	10.93 61	6.48 65	8.51 50	4.96 45	0.02 38	0.15 98	0.00 1	0.00 1	0.17 56	0.17 61
Konstantinos Batsos, Changjiang Cai, Philippos Mordohai: CBMV: A Coalesced Bidirectional Matching Volume for Disparity Estimation. Computer Vision and Pattern Recognition (CVPR) 2018
ADCP+	8.09 92	1.79 96	14.50 111	1.54 25	4.28 110	16.57 109	5.20 90	12.80 73	11.20 67	12.83 88	17.07 83	11.02 62	10.80 92	17.59 115	23.18 126	0.03 49	0.05 67	0.01 59	0.18 97	0.39 92	0.81 109

DN-CSS_ROB	2.69 9	1.40 77	5.34 50	2.31 69	0.75 27	3.14 15	0.06 2	6.11 5	3.87 3	5.34 16	12.18 57	2.34 6	1.22 6	7.84 37	1.48 6	0.03 49	0.00 1	0.00 1	0.00 1	0.35 85	0.03 17

ac_64	3.70 29	0.47 12	3.50 24	2.87 88	3.50 100	5.91 48	0.72 34	10.17 39	4.34 5	8.16 45	5.80 11	10.82 59	5.64 59	7.08 26	4.73 43	0.03 49	0.00 1	0.00 1	0.00 1	0.08 31	0.14 56

hitnet-ft	3.66 26	0.65 30	2.79 11	0.89 6	0.72 26	4.25 27	0.55 19	9.11 21	7.52 37	7.14 33	10.19 39	12.49 70	4.70 46	7.44 29	4.41 39	0.03 49	0.00 1	0.06 91	0.00 1	0.20 61	0.04 27

FAT-Stereo	6.78 80	0.68 31	6.80 77	2.30 67	1.77 64	5.63 46	4.20 78	18.79 112	18.62 99	10.53 70	17.15 84	18.52 100	13.74 106	8.86 57	7.38 64	0.03 49	0.15 98	0.01 59	0.07 85	0.12 43	0.26 77

FC-DCNN	10.72 114	0.52 15	4.27 34	1.88 41	1.63 56	17.18 113	5.29 91	18.20 110	19.69 106	28.50 125	34.51 127	34.03 131	21.48 125	15.89 109	11.15 101	0.03 49	0.01 39	0.02 71	0.01 57	0.07 29	0.09 40

edge stereo	6.76 77	1.01 52	6.76 76	2.20 61	2.45 80	6.41 53	2.45 58	14.84 93	11.98 74	15.29 97	18.31 90	22.02 113	12.56 98	10.82 76	7.49 65	0.03 49	0.06 73	0.11 100	0.03 70	0.30 78	0.14 56

FADNet-RVC	3.98 39	1.84 98	12.48 104	1.69 35	0.44 16	4.33 28	1.31 47	11.84 61	7.15 30	3.53 4	3.50 4	10.63 58	4.43 41	9.12 60	6.25 55	0.03 49	0.10 87	0.00 1	0.03 70	0.60 100	0.25 74

DRN-Test	5.87 63	0.98 51	5.89 60	2.69 83	3.65 105	12.37 96	3.35 67	20.07 124	10.20 62	11.93 80	12.31 60	11.06 63	5.31 55	7.89 38	9.05 84	0.04 57	0.05 67	0.04 86	0.04 79	0.18 59	0.25 74

ETE_ROB	5.80 62	1.77 95	6.33 66	1.44 22	0.78 31	6.43 54	6.90 105	12.53 71	8.08 45	12.93 89	14.89 74	21.13 112	5.87 61	9.83 67	6.57 60	0.04 57	0.01 39	0.00 1	0.02 61	0.08 31	0.33 82

EDNetEfficientorigin	17.87 134	14.00 137	99.53 144	1.64 32	0.92 38	10.55 86	6.26 99	13.21 78	34.84 128	38.20 133	36.91 130	27.47 122	26.32 133	14.71 105	25.69 130	0.04 57	0.06 73	0.14 105	0.52 121	3.16 131	3.27 132

LALA_ROB	6.58 74	1.80 97	6.25 64	1.26 14	0.94 41	10.08 81	9.02 112	16.00 102	11.51 71	12.74 86	13.02 65	24.77 116	5.25 54	10.56 74	8.02 71	0.04 57	0.05 67	0.00 1	0.02 61	0.10 36	0.25 74

psmorigin	7.34 85	1.58 86	18.31 117	2.35 71	0.87 33	6.72 57	1.70 54	10.63 48	7.14 29	19.77 109	22.71 98	20.13 108	11.34 94	12.96 92	9.49 87	0.04 57	0.17 102	0.06 91	0.17 96	0.21 68	0.50 94

FCDSN-DC	10.24 111	0.56 20	3.49 23	1.96 49	1.29 49	16.90 112	4.59 87	17.16 108	19.10 102	24.64 121	32.46 123	33.82 130	22.14 127	15.93 110	10.45 94	0.04 57	0.00 1	0.00 1	0.00 1	0.05 22	0.21 65
Dominik Hirner, Friedrich Fraundorfer: FCDSN-DC: An accurate but lightweight end-to-end trainable neural network for stereo estimation with depth completion.
BEATNet_4x	3.24 15	1.27 71	5.89 60	1.56 26	0.10 1	5.26 40	1.07 41	10.08 34	5.50 12	6.89 31	7.73 23	4.53 10	4.13 36	5.05 14	5.27 47	0.04 57	0.05 67	0.00 1	0.00 1	0.23 75	0.23 68

NaN_ROB	6.00 65	1.24 68	6.29 65	1.34 16	1.68 60	9.60 78	10.31 120	15.09 94	15.79 88	12.62 84	8.95 33	11.67 66	5.83 60	11.78 84	6.41 58	0.05 64	0.13 95	0.08 95	0.20 99	0.22 70	0.79 108

LE_ROB	16.73 131	1.28 74	11.61 99	3.72 111	1.65 57	16.67 110	9.17 113	14.39 89	55.91 140	63.81 140	40.86 135	35.94 133	37.73 139	14.24 101	26.87 132	0.05 64	0.10 87	0.13 104	0.22 100	0.12 43	0.15 59

DMCA-RVC	3.17 14	1.26 70	6.39 69	2.10 56	1.51 53	3.02 14	0.56 21	7.84 11	5.06 7	9.23 57	5.19 8	8.46 41	2.72 18	6.31 19	3.22 24	0.06 66	0.11 89	0.09 98	0.02 61	0.16 54	0.08 36

pmcnn	7.72 88	1.27 71	9.42 91	2.91 89	3.14 90	9.44 74	6.23 97	12.56 72	16.51 92	14.53 94	24.08 102	27.44 121	8.49 79	9.32 64	8.44 78	0.06 66	0.08 82	0.00 1	0.00 1	0.30 78	0.15 59

DeepPruner_ROB	3.52 23	1.14 64	4.06 32	1.12 12	1.65 57	3.65 17	0.83 38	13.96 85	4.47 6	7.80 39	10.84 45	7.05 29	2.16 13	8.14 46	3.08 21	0.07 68	0.03 58	0.00 1	0.01 57	0.32 81	0.06 31

SANet	10.64 113	1.86 99	10.91 94	1.76 36	0.71 25	14.62 105	9.23 114	19.18 115	37.14 132	19.22 105	27.96 110	25.86 119	19.11 118	13.02 93	10.63 95	0.08 69	0.06 73	0.03 79	0.02 61	0.62 101	0.81 109

CSAN	7.62 86	1.60 88	6.56 73	1.83 38	0.66 22	12.40 97	10.52 122	14.45 90	21.32 110	14.19 93	15.98 80	17.84 94	13.02 103	12.32 87	8.38 77	0.09 70	0.07 80	0.03 79	0.04 79	0.33 82	0.67 106

ccs	3.04 12	0.39 9	3.08 17	1.78 37	0.52 19	2.04 2	0.50 17	13.09 76	13.71 80	3.54 5	5.36 10	5.50 14	2.45 14	4.81 13	2.88 17	0.09 70	0.08 82	0.12 101	0.10 91	0.20 61	0.50 94

ADCS	13.02 118	4.93 124	28.38 129	3.17 100	2.67 85	13.61 103	10.83 123	18.70 111	33.46 126	22.59 113	24.78 105	19.59 105	18.51 116	23.40 127	32.16 138	0.10 72	0.19 105	0.37 123	0.18 97	1.26 118	1.46 122

SGM-Forest	4.96 56	0.32 1	2.84 12	1.21 13	0.64 21	10.23 84	6.64 104	11.55 58	10.98 65	10.94 75	13.59 67	11.65 65	4.30 40	8.94 58	4.63 42	0.11 73	0.04 63	0.00 1	0.00 1	0.05 22	0.46 91
Johannes L. Schönberger, Sudipta Sinha, Marc Pollefeys: Learning to Fuse Proposals from Multiple Scanline Optimizations in Semi-Global Matching. ECCV 2018
MeshStereo	11.52 116	1.52 83	4.55 40	1.89 42	1.46 51	19.87 118	5.11 89	20.66 126	15.91 89	32.67 130	34.51 127	39.34 136	21.15 123	18.74 118	12.10 103	0.11 73	0.06 73	0.01 59	0.00 1	0.45 95	0.22 67
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
SAMSARA	14.63 122	2.74 110	12.38 102	12.65 137	6.74 130	36.50 134	72.93 142	19.36 117	23.77 114	16.20 100	13.04 66	29.21 123	12.78 100	16.98 113	15.21 111	0.11 73	0.26 112	0.03 79	0.14 94	0.76 106	0.77 107

S-Stereo	6.63 75	0.84 40	9.67 92	3.15 99	3.48 99	6.49 55	6.22 96	12.99 75	22.84 111	9.48 59	15.51 77	12.00 67	8.43 78	8.04 41	10.70 96	0.12 76	0.17 102	0.00 1	0.38 113	0.13 47	1.92 125

CC-Net-ROB	3.84 34	1.07 57	5.23 48	2.65 81	2.96 89	4.22 26	0.69 32	10.43 44	7.72 39	8.78 50	8.29 27	9.61 49	4.02 35	7.16 28	3.65 33	0.13 77	0.03 58	0.02 71	0.03 70	0.05 22	0.03 17

SuperB	8.10 93	3.15 116	24.67 126	2.65 81	1.23 46	9.88 80	4.29 79	10.18 40	30.07 122	11.53 78	12.18 57	6.12 21	6.65 69	10.50 73	14.47 110	0.14 78	0.11 89	0.35 121	0.25 103	13.06 138	0.48 92

NLCA_NET_v2_RVC	3.84 34	1.06 56	5.23 48	2.72 85	3.27 92	4.36 29	0.61 24	10.71 49	7.56 38	8.75 49	7.89 24	9.86 51	3.90 33	7.15 27	3.44 30	0.14 78	0.02 51	0.02 71	0.03 70	0.04 18	0.03 17
Zhibo Rao, Mingyi He, Yuchao Dai, Zhidong Zhu, Bo Li, and Renjie He.: NLCA-Net: A non-local context attention network for stereo matching.
ADCRef	7.27 84	1.38 76	16.37 113	2.52 76	3.30 94	11.63 94	3.16 64	10.80 51	9.35 54	13.03 90	25.27 108	8.17 39	8.92 83	8.06 43	21.81 123	0.15 80	0.08 82	0.16 107	0.34 111	0.38 88	0.58 100

EDNetEfficient	13.82 120	6.19 127	54.79 136	2.83 86	3.60 102	4.86 37	8.57 110	8.61 18	32.90 125	40.77 136	30.59 117	18.33 99	21.87 126	11.38 82	25.36 129	0.15 80	0.23 109	0.06 91	1.03 128	2.54 128	1.79 123

RASNet	4.52 55	0.61 25	4.42 36	3.42 105	4.68 120	4.58 34	0.99 40	9.54 28	8.01 44	5.28 15	11.42 51	10.34 56	8.88 82	9.28 63	8.68 80	0.15 80	0.00 1	0.00 1	0.00 1	0.03 14	0.04 27

LSM	14.01 121	5.95 126	33.49 131	6.78 132	43.61 140	10.22 83	9.98 119	15.16 97	22.93 112	23.07 117	32.34 122	18.52 100	12.67 99	15.45 108	11.10 100	0.16 83	0.51 125	0.09 98	0.32 109	1.08 114	16.85 140

FADNet-RVC-Resample	3.79 32	1.62 89	12.06 101	1.43 21	0.66 22	5.94 49	2.41 57	10.18 40	8.58 51	6.28 24	4.22 5	5.33 13	4.80 50	7.71 34	3.19 23	0.17 84	0.21 106	0.17 108	0.12 93	0.41 93	0.29 81

AnyNet_C32	10.98 115	5.58 125	22.79 123	4.16 117	5.83 127	15.64 106	14.30 128	13.18 77	17.15 96	16.44 101	20.52 93	14.68 82	13.44 104	22.46 124	30.08 136	0.17 84	0.26 112	0.36 122	0.36 112	1.23 117	0.91 112

PASM	7.90 90	4.22 122	21.97 122	3.25 101	3.29 93	5.39 42	6.57 103	10.57 46	19.09 101	12.77 87	13.92 69	18.11 97	9.51 87	13.79 99	10.77 99	0.19 86	0.45 121	0.29 117	1.08 129	1.49 122	1.19 118

DANet	6.02 66	1.23 67	8.45 88	3.86 114	3.94 109	7.64 67	1.34 48	9.51 27	7.00 27	13.39 91	15.53 78	15.99 88	7.02 71	12.14 85	12.37 104	0.19 86	0.12 94	0.02 71	0.03 70	0.13 47	0.56 99

FADNet	4.23 47	1.65 91	11.75 100	1.64 32	0.80 32	4.80 35	0.77 37	13.76 84	11.65 72	3.97 6	5.24 9	9.62 50	5.14 53	8.40 49	3.78 34	0.21 88	0.04 63	0.07 94	0.05 81	1.14 115	0.10 47

psm_up	3.85 36	0.82 38	3.01 14	2.00 51	2.58 84	5.28 41	3.34 66	10.85 53	6.33 21	7.27 35	7.32 20	10.15 52	9.02 84	6.19 17	2.51 14	0.22 89	0.01 39	0.00 1	0.00 1	0.12 43	0.03 17

PVD	15.44 124	2.93 112	14.67 112	4.21 120	3.39 98	17.43 114	4.16 76	27.84 136	48.84 139	31.02 128	43.54 136	29.76 125	30.81 136	25.97 130	21.40 120	0.23 90	0.41 117	0.04 86	0.33 110	0.41 93	1.33 119

DeepPrunerF	6.75 76	2.69 109	23.31 124	3.68 109	7.16 131	3.78 21	4.29 79	13.42 80	20.13 108	8.13 44	10.46 41	7.18 32	8.06 76	11.10 80	9.44 86	0.24 91	0.15 98	0.29 117	0.42 116	0.66 103	0.45 89

aanetorigin	8.72 100	3.29 119	27.55 128	1.95 48	3.84 107	4.93 39	5.39 92	5.49 3	10.05 61	27.85 124	31.47 119	15.80 85	12.78 100	9.04 59	11.98 102	0.25 92	0.22 107	0.22 114	0.25 103	1.28 119	0.84 111

SGM_RVC	10.08 108	0.60 23	3.42 22	2.30 67	0.32 11	19.41 116	6.33 100	18.95 114	14.64 82	25.14 122	24.32 104	33.34 129	18.79 117	19.86 120	12.55 106	0.25 92	0.26 112	0.22 114	0.24 102	0.34 84	0.40 86
Heiko Hirschmueller: Stereo processing by semiglobal matching and mutual information. TPAMI 2008, Volume 30(2), pp. 328-341
SGM-ForestM	16.99 132	1.08 59	5.74 57	2.12 57	0.75 27	31.63 132	12.21 125	27.80 135	32.25 124	37.88 132	39.99 133	52.96 139	35.20 138	33.60 135	24.47 127	0.26 94	0.39 116	0.31 119	0.39 114	0.26 76	0.53 98

stereogan	7.69 87	0.88 42	7.08 80	3.49 107	3.93 108	18.98 115	3.23 65	16.52 104	19.58 105	9.93 66	18.92 91	20.50 110	9.04 85	14.07 100	6.14 54	0.26 94	0.04 63	0.21 113	0.03 70	0.63 102	0.33 82

ADCL	10.16 110	2.11 103	19.36 119	1.92 44	1.88 67	22.23 120	8.91 111	14.04 86	23.56 113	14.62 95	26.19 109	12.75 71	13.59 105	16.06 111	22.95 125	0.26 94	0.18 104	0.75 127	0.65 122	0.69 104	0.58 100

ADCPNet	9.54 104	2.39 107	31.46 130	2.09 55	1.60 55	16.71 111	6.39 102	12.11 64	11.45 69	13.53 92	21.45 95	19.41 104	10.94 93	14.38 102	21.54 121	0.27 97	1.16 129	0.39 124	1.49 133	0.58 98	1.45 121

AnyNet_C01	16.12 127	10.81 136	59.36 137	4.42 121	2.49 82	30.06 130	15.15 134	17.51 109	16.51 92	17.88 102	37.69 131	24.04 115	17.54 112	29.60 133	33.29 139	0.28 98	0.38 115	0.43 125	0.42 116	2.57 129	1.98 126

RTSC	9.15 101	3.00 113	13.57 109	3.72 111	1.76 63	11.82 95	0.46 14	16.95 106	36.83 131	15.80 98	15.53 78	12.91 72	7.46 74	20.01 122	21.76 122	0.31 99	0.13 95	0.01 59	0.08 87	0.57 97	0.41 88

SHD	9.61 105	2.60 108	12.46 103	3.69 110	3.54 101	9.47 75	1.25 44	20.16 125	37.84 134	18.19 103	21.24 94	16.96 91	12.83 102	14.47 104	16.05 112	0.32 100	0.13 95	0.01 59	0.08 87	0.38 88	0.48 92

MSMD_ROB	9.28 103	1.09 60	4.65 42	1.58 27	0.39 13	16.52 108	4.41 82	13.60 81	14.87 84	22.34 112	39.89 132	25.67 118	20.71 122	12.42 88	6.98 61	0.34 101	0.03 58	0.00 1	0.00 1	0.05 22	0.09 40

FADNet_RVC	3.91 37	1.67 92	12.95 107	0.96 9	0.75 27	5.71 47	0.54 18	10.83 52	6.60 24	3.46 3	8.09 25	4.10 9	3.40 24	9.43 65	6.33 56	0.36 102	0.44 120	0.17 108	0.46 120	0.91 110	0.95 113

GANet	6.22 69	1.07 57	4.07 33	2.27 65	0.89 34	9.19 72	9.52 115	12.02 63	8.13 46	10.72 72	29.09 114	13.86 75	7.52 75	11.00 79	4.39 38	0.36 102	0.00 1	0.02 71	0.02 61	0.12 43	0.08 36

XQC	8.43 98	3.58 120	16.40 114	2.92 91	2.17 72	13.22 101	3.60 70	14.64 92	25.86 116	11.87 79	12.04 55	15.06 83	10.67 91	15.24 107	19.41 115	0.39 104	0.08 82	0.05 89	0.07 85	0.84 109	0.45 89

CVANet_RVC	4.16 43	1.16 65	3.60 26	1.94 47	1.46 51	3.92 24	4.68 88	10.89 54	8.34 48	7.58 37	10.84 45	10.27 55	6.62 68	8.56 51	2.69 15	0.39 104	0.00 1	0.00 1	0.01 57	0.21 68	0.09 40

RTSA	18.87 135	9.32 134	86.48 139	4.95 124	6.10 128	42.08 137	14.70 130	15.49 99	41.06 136	22.65 114	32.32 120	13.77 73	19.54 119	37.98 136	28.96 133	0.41 106	0.23 109	0.00 1	0.02 61	0.91 110	0.50 94

GANetREF_RVC	6.56 73	2.89 111	7.58 85	3.41 104	0.40 14	12.96 99	9.58 116	15.09 94	17.25 97	10.33 68	10.62 44	12.27 69	8.16 77	12.21 86	4.53 41	0.41 106	0.00 1	0.00 1	0.02 61	3.12 130	0.39 85
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
RTS	18.87 135	9.32 134	86.48 139	4.95 124	6.10 128	42.08 137	14.70 130	15.49 99	41.06 136	22.65 114	32.32 120	13.77 73	19.54 119	37.98 136	28.96 133	0.41 106	0.23 109	0.00 1	0.02 61	0.91 110	0.50 94

PWC_ROB	8.24 94	3.13 114	12.74 105	2.43 74	4.43 114	7.51 66	1.22 43	16.63 105	19.24 103	16.08 99	28.29 111	13.99 78	10.16 89	13.63 98	14.06 109	0.42 109	0.00 1	0.05 89	0.00 1	0.59 99	0.27 79

ADCMid	10.24 111	3.13 114	20.70 120	2.21 62	2.39 78	11.23 92	6.19 95	14.17 88	11.19 66	23.20 118	22.25 96	17.89 96	19.54 119	18.51 117	26.21 131	0.45 110	0.42 119	1.10 131	1.29 130	1.56 124	1.18 117

BEATNet-Init1	23.31 138	9.03 133	41.67 133	4.17 118	2.53 83	45.68 139	19.47 137	33.43 137	38.45 135	47.59 139	49.10 139	59.31 140	41.80 141	38.35 138	29.21 135	0.47 111	0.50 124	0.81 128	0.66 123	2.10 127	1.86 124

STTStereo_v2	8.41 96	1.54 84	10.97 96	5.73 127	3.60 102	26.19 125	4.41 82	10.10 35	7.42 33	19.71 107	24.99 106	14.38 80	15.83 109	10.99 77	9.53 88	0.50 112	0.46 122	0.19 111	0.25 103	0.80 107	0.66 104

G-Net	8.41 96	1.54 84	10.97 96	5.73 127	3.60 102	26.19 125	4.41 82	10.10 35	7.42 33	19.71 107	24.99 106	14.38 80	15.83 109	10.99 77	9.53 88	0.50 112	0.46 122	0.19 111	0.25 103	0.80 107	0.66 104

TDLM	4.11 41	1.11 62	3.54 25	1.62 30	1.04 43	3.91 23	7.41 109	10.60 47	10.67 64	6.38 25	12.59 62	5.95 20	4.77 48	8.79 55	3.04 19	0.58 114	0.00 1	0.01 59	0.00 1	0.19 60	0.12 51

AF-Net	7.78 89	1.44 79	6.68 75	3.37 103	4.50 118	8.61 69	2.69 59	17.07 107	20.17 109	9.52 60	24.02 101	20.31 109	14.59 108	11.58 83	9.84 92	0.61 115	0.00 1	0.12 101	0.00 1	0.38 88	0.12 51

FBW_ROB	8.50 99	1.03 54	7.98 87	1.93 46	1.28 48	13.10 100	6.23 97	22.50 128	18.98 100	18.82 104	14.91 75	19.06 103	10.04 88	18.41 116	9.83 91	0.62 116	0.22 107	1.82 135	0.82 126	0.99 113	1.36 120

RGC	6.88 82	2.23 105	6.13 63	4.05 116	4.73 122	8.94 71	2.78 61	15.19 98	11.74 73	11.13 76	19.34 92	17.86 95	10.42 90	13.02 93	8.03 73	0.73 117	0.01 39	0.24 116	0.41 115	0.31 80	0.38 84

DPSNet	10.14 109	1.88 100	16.82 115	1.85 39	1.73 62	24.84 124	17.20 136	19.92 123	27.41 118	12.23 83	13.62 68	16.52 89	18.35 114	14.42 103	12.50 105	0.78 118	0.54 126	0.08 95	0.25 103	1.18 116	0.59 103

Nwc_Net	6.97 83	1.25 69	6.63 74	3.82 113	3.37 95	10.83 90	1.67 51	19.56 122	11.35 68	8.36 46	23.62 100	17.19 93	11.44 95	11.21 81	8.08 74	0.80 119	0.00 1	0.00 1	0.02 61	0.13 47	0.09 40

WCMA_ROB	9.21 102	0.87 41	7.37 82	2.54 78	2.13 70	13.59 102	5.80 93	11.64 59	14.01 81	24.43 120	32.99 124	27.09 120	18.02 113	12.51 89	9.85 93	0.81 120	0.07 80	0.01 59	0.01 57	0.16 54	0.23 68

ELAS	16.72 130	2.14 104	9.23 90	4.92 123	4.53 119	32.66 133	15.11 133	27.40 134	28.68 121	40.27 135	44.90 137	38.33 135	30.50 135	26.44 131	21.94 124	0.88 121	1.23 131	0.67 126	0.89 127	1.49 122	2.18 130
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
MANE	19.47 137	1.27 71	5.07 47	4.69 122	5.55 125	30.49 131	9.94 118	34.01 138	37.27 133	44.13 137	51.57 141	52.51 138	40.41 140	33.58 134	24.81 128	0.89 122	0.86 127	1.11 132	9.72 139	0.38 88	1.06 115

RP	6.84 81	1.29 75	5.53 54	3.92 115	5.18 123	6.32 51	3.53 69	11.73 60	15.31 86	9.54 61	22.38 97	18.25 98	14.47 107	10.11 70	7.49 65	0.91 123	0.01 39	0.12 101	0.15 95	0.33 82	0.19 63

AANet_RVC	5.01 58	1.74 93	6.38 68	1.96 49	1.29 49	2.26 3	1.69 52	10.07 33	18.53 98	7.88 40	18.15 89	8.49 44	2.70 17	10.59 75	7.04 62	0.96 124	0.15 98	0.02 71	0.00 1	0.13 47	0.12 51

ELAS_RVC	16.54 129	2.26 106	10.09 93	5.50 126	4.46 116	28.28 129	16.72 135	25.55 132	33.54 127	40.19 134	40.30 134	36.68 134	30.03 134	29.40 132	20.61 118	0.98 125	1.21 130	0.86 129	0.70 124	1.39 121	2.16 128
A. Geiger, M. Roser, R. Urtasun: Efficient large-scale stereo matching. ACCV 2010
MFN_U_SF_RVC	12.94 117	3.66 121	25.81 127	3.61 108	2.26 75	22.77 121	4.55 86	27.10 133	20.06 107	23.90 119	28.99 113	30.53 126	16.98 111	19.92 121	20.26 117	1.24 126	1.07 128	0.98 130	1.33 131	1.80 125	2.04 127

MFN_U_SF_DS_RVC	9.78 107	4.27 123	14.47 110	2.29 66	2.85 88	23.40 123	13.62 127	13.60 81	19.54 104	19.42 106	24.27 103	16.74 90	8.59 80	17.05 114	7.98 70	1.25 127	1.68 133	0.17 108	2.63 135	0.72 105	1.04 114

NCC-stereo	6.77 78	1.49 81	6.48 70	2.92 91	4.40 111	7.43 63	3.61 71	19.52 120	13.29 77	8.39 47	16.91 81	15.96 86	12.13 96	12.85 90	7.70 67	1.47 128	0.11 89	0.01 59	0.42 116	0.14 52	0.24 71

Abc-Net	6.77 78	1.49 81	6.48 70	2.92 91	4.40 111	7.43 63	3.61 71	19.52 120	13.29 77	8.39 47	16.91 81	15.96 86	12.13 96	12.85 90	7.70 67	1.47 128	0.11 89	0.01 59	0.42 116	0.14 52	0.24 71
Xing Li, Yangyu Fan, Guoyun Lv, and Haoyue Ma: Area-based Correlation and Non-local Attention Network for Stereo Matching. The Visual Computer
MADNet+	27.07 139	33.84 139	90.97 141	20.14 138	7.47 135	48.43 140	47.10 139	35.43 139	36.46 130	20.11 111	30.05 116	25.29 117	35.08 137	45.50 140	50.28 140	2.13 130	2.00 134	1.19 134	0.76 125	4.71 133	4.43 133

MFMNet_re	13.29 119	8.60 132	18.29 116	9.75 134	7.25 133	19.65 117	14.84 132	20.71 127	30.72 123	23.03 116	28.77 112	18.85 102	26.09 132	13.55 97	9.82 90	2.44 131	1.35 132	0.34 120	0.23 101	4.78 134	6.69 135

MSC_U_SF_DS_RVC	16.41 128	6.93 130	21.83 121	5.94 130	2.81 87	38.71 135	14.59 129	24.55 131	34.87 129	33.66 131	34.35 126	29.24 124	24.20 129	22.59 125	17.95 114	2.52 132	2.81 135	1.17 133	1.51 134	5.89 136	2.16 128

PWCDC_ROB	7.92 91	3.17 118	7.48 84	5.73 127	4.40 111	10.45 85	0.35 11	14.52 91	28.19 119	10.36 69	31.27 118	7.04 27	9.14 86	13.22 95	8.78 82	2.74 133	0.02 51	0.00 1	0.00 1	1.31 120	0.17 61

PMTNet	1.41 2	0.37 8	2.06 4	0.26 3	0.15 4	2.94 13	0.34 10	3.39 1	5.33 11	1.00 2	1.27 1	0.28 1	0.93 5	4.04 9	0.74 4	5.00 134	0.00 1	0.00 1	0.00 1	0.03 14	0.01 2

DispFullNet	17.47 133	26.01 138	33.98 132	22.58 139	20.86 139	13.84 104	1.28 45	16.50 103	26.27 117	19.97 110	17.17 85	20.52 111	18.49 115	22.86 126	10.76 98	5.13 135	2.83 136	30.72 140	7.72 138	20.86 139	11.01 139

PDISCO_ROB	9.62 106	1.99 102	11.51 98	9.88 135	9.61 136	21.48 119	3.83 74	19.33 116	28.49 120	11.27 77	14.17 70	19.92 107	5.02 52	16.35 112	9.18 85	5.28 136	0.41 117	0.14 105	0.09 90	2.05 126	2.36 131

NVStereoNet_ROB	16.04 126	6.75 129	12.90 106	6.37 131	7.42 134	12.89 98	9.74 117	22.78 129	25.12 115	30.32 126	46.19 138	34.37 132	25.38 130	21.48 123	21.38 119	5.94 137	3.10 137	6.07 138	10.09 140	4.01 132	8.54 138
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-STEREO	15.04 123	6.23 128	13.21 108	11.34 136	11.65 138	23.30 122	7.15 107	24.16 130	15.65 87	31.78 129	29.19 115	31.62 127	21.32 124	24.62 128	19.50 116	7.59 138	4.19 139	3.22 136	1.48 132	6.99 137	6.54 134
K. Yamaguchi, D. McAllester, R. Urtasun: Efficient Joint Segmentation, Occlusion Labeling, Stereo and Flow Estimation. ECCV 2014
SGM+DAISY	15.62 125	7.26 131	19.28 118	8.94 133	10.11 137	26.25 127	10.49 121	19.36 117	14.65 83	30.64 127	33.59 125	33.00 128	22.32 128	24.96 129	16.42 113	7.90 139	6.25 140	4.51 137	3.37 136	5.86 135	7.20 136

PWCK	30.53 140	44.32 140	47.25 135	29.76 140	7.23 132	40.78 136	27.10 138	44.73 140	44.32 138	47.31 138	36.37 129	47.16 137	26.05 131	41.26 139	31.87 137	21.83 140	4.03 138	29.50 139	4.67 137	27.17 140	7.80 137

DPSimNet_ROB	53.45 141	64.73 141	44.39 134	53.97 141	45.39 141	53.66 141	54.83 140	55.15 141	57.87 141	64.16 141	50.83 140	63.40 141	53.34 142	46.45 141	65.81 141	63.13 141	26.54 141	57.94 141	51.11 141	45.52 141	50.69 141

MADNet++	82.84 142	82.38 143	73.57 138	87.72 142	82.97 142	93.14 142	69.15 141	86.42 142	82.50 142	93.46 142	86.70 142	86.28 142	80.92 143	88.34 142	88.84 142	86.83 142	84.17 142	72.64 142	68.92 142	80.47 142	81.42 142

DPSM_ROB	99.95 147	100.00 146	100.00 147	99.76 144	100.00 144	100.00 145	100.00 145	100.00 146	100.00 144	100.00 146	100.00 144	100.00 146	100.00 145	100.00 146	100.00 147	99.21 143	100.00 144	100.00 143	100.00 144	99.99 144	99.95 144

DPSM	99.95 147	100.00 146	100.00 147	99.76 144	100.00 144	100.00 145	100.00 145	100.00 146	100.00 144	100.00 146	100.00 144	100.00 146	100.00 145	100.00 146	100.00 147	99.21 143	100.00 144	100.00 143	100.00 144	99.99 144	99.95 144

MEDIAN_ROB	98.41 143	99.70 144	99.30 143	97.09 143	97.02 143	96.89 143	95.77 144	97.66 143	97.28 143	98.79 145	98.94 143	99.18 143	98.14 144	96.89 143	96.88 143	99.96 145	99.16 143	100.00 143	99.99 143	99.69 143	99.88 143

DGTPSM_ROB	99.90 145	100.00 146	99.99 145	99.99 146	100.00 144	100.00 145	100.00 145	99.97 144	100.00 144	98.35 143	100.00 144	99.84 144	100.00 145	99.98 144	99.99 144	99.99 146	100.00 144	100.00 143	100.00 144	100.00 146	100.00 147

AVERAGE_ROB	99.62 144	99.95 145	98.81 142	100.00 148	100.00 144	98.08 144	95.47 143	100.00 146	100.00 144	100.00 146	100.00 144	100.00 146	100.00 145	100.00 146	99.99 144	100.00 147	100.00 144	100.00 143	100.00 144	100.00 146	100.00 147

DPSMNet_ROB	99.91 146	100.00 146	99.99 145	99.99 146	100.00 144	100.00 145	100.00 145	99.98 145	100.00 144	98.35 143	100.00 144	99.84 144	100.00 145	99.98 144	99.99 144	100.00 147	100.00 144	100.00 143	100.00 144	100.00 146	100.00 147

LSM0	100.00 149	100.00 146	100.00 147	100.00 148	100.00 144	100.00 145	100.00 145	100.00 146	100.00 144	100.00 146	100.00 144	100.00 146	100.00 145	100.00 146	100.00 147	100.00 147	100.00 144	100.00 143	100.00 144	100.00 146	99.99 146

MSMDNet													1.26 7

ASD4		81.57 142

Low-res two-view benchmark