High-res multi-view results

This table lists the benchmark results for the high-res multi-view scenario. The following metrics are evaluated:

Accuracy [%]: The fraction of the reconstruction which is closer to the ground truth than the evaluation threshold distance (*). Larger is better.
Completeness [%]: The fraction of the ground truth which is closer to the reconstruction than the evaluation threshold distance (*). Larger is better.
F₁ score [%]: The harmonic mean of accuracy and completeness, used to rank methods based on both metrics. Larger is better.
Time [s]: The runtime of the method. Smaller is better.

(*) For exact definitions, detailing how potentially incomplete ground truth is taken into account, see our paper.

The datasets are grouped into different categories, and result averages are computed for a category and method if results of the method are available for all datasets within the category. Note that the category "all" includes both the high-res multi-view and the low-res many-view scenarios.

Methods with suffix _ROB may participate in the Robust Vision Challenge.

Click a dataset result cell to show a visualization of the reconstruction. For training datasets, ground truth and accuracy / completeness visualizations are also available. The visualizations may not work with mobile browsers.

Method	all	high-res multi-view	indoor	outdoor	botani.	boulde.	bridge	door	exhibi.	lectur.	living.	lounge	observ.	old co.	statue	terrac.

SteuartSystems-ACMMP		99.54 1	99.42 1	99.87 1	99.82 9	99.81 1	99.59 27	99.51 47	98.62 14	99.79 5	99.83 26	99.28 1	99.97 5	98.48 44	99.90 12	99.84 5
Skip Steuart: Steuart Systems R&D Blog.
HPM-MVS_fast		99.51 2	99.40 3	99.85 2	99.91 1	99.79 2	99.76 13	99.56 26	97.72 48	99.76 9	99.75 46	99.13 2	99.92 22	99.07 10	99.92 4	99.85 3

DeepC-MVS_fast	98.69 1	99.49 3	99.39 4	99.77 12	99.63 38	99.59 14	99.36 69	99.46 66	99.07 7	99.79 5	99.82 30	98.85 5	99.92 22	98.68 34	99.87 20	99.82 9
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
DELS-MVS		99.48 4	99.42 1	99.65 20	99.72 25	99.40 25	99.05 105	99.66 15	99.14 5	99.57 26	99.80 37	98.46 17	99.94 14	99.57 1	99.84 29	99.60 45

ACMMP		99.45 5	99.32 7	99.82 6	99.89 3	99.67 9	99.62 21	99.69 13	98.12 27	99.63 19	99.84 24	98.73 11	99.96 8	98.55 40	99.83 32	99.81 10
Qingshan Xu, Weihang Kong, Wenbing Tao, Marc Pollefeys: Multi-Scale Geometric Consistency Guided and Planar Prior Assisted Multi-View Stereo. IEEE Transactions on Pattern Analysis and Machine Intelligence
CP-MVS		99.45 5	99.32 7	99.85 2	99.83 8	99.75 4	99.69 19	99.52 42	98.07 32	99.53 31	99.63 66	98.93 3	99.97 5	98.74 30	99.91 6	99.83 8

HPM-MVS		99.42 7	99.28 11	99.83 5	99.90 2	99.72 6	99.81 7	99.54 36	97.59 52	99.68 12	99.63 66	98.91 4	99.94 14	98.58 37	99.91 6	99.84 5

MVS_111021_LR		99.41 8	99.33 6	99.65 20	99.77 15	99.51 20	98.94 116	99.85 4	98.82 12	99.65 18	99.74 48	98.51 15	99.80 57	98.83 28	99.89 17	99.64 42

MVS_111021_HR		99.41 8	99.32 7	99.66 17	99.72 25	99.47 21	98.95 115	99.85 4	98.82 12	99.54 30	99.73 49	98.51 15	99.74 66	98.91 20	99.88 18	99.77 18

PVSNet_Blended_VisFu		99.36 10	99.28 11	99.61 25	99.86 5	99.07 38	99.47 47	99.93 1	97.66 51	99.71 10	99.86 18	97.73 32	99.96 8	99.47 2	99.82 33	99.79 16

NCCC		99.34 11	99.19 13	99.79 10	99.61 41	99.65 13	99.30 78	99.48 56	98.86 11	99.21 61	99.63 66	98.72 12	99.90 31	98.25 53	99.63 46	99.80 14

CSCG		99.32 12	99.32 7	99.32 50	99.85 6	98.29 66	99.71 17	99.66 15	98.11 28	99.41 44	99.80 37	98.37 20	99.96 8	98.99 15	99.96 2	99.72 26

DeepC-MVS	98.35 2	99.30 13	99.19 13	99.64 23	99.82 9	99.23 31	99.62 21	99.55 30	98.94 10	99.63 19	99.95 1	95.82 59	99.94 14	99.37 3	99.97 1	99.73 22
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
PHI-MVS		99.30 13	99.17 15	99.70 16	99.56 48	99.52 19	99.58 29	99.80 6	97.12 79	99.62 22	99.73 49	98.58 14	99.90 31	98.61 36	99.91 6	99.68 32

LS3D		99.27 15	99.12 16	99.74 13	99.18 90	99.75 4	99.56 35	99.57 24	98.45 19	99.49 36	99.85 21	97.77 31	99.94 14	98.33 50	99.84 29	99.52 57

APD-MVS		99.27 15	99.08 18	99.84 4	99.75 20	99.79 2	99.50 44	99.50 51	97.16 76	99.77 8	99.82 30	98.78 6	99.94 14	97.56 80	99.86 22	99.80 14

3Dnovator	97.25 9	99.24 17	99.05 20	99.81 8	99.12 100	99.66 11	99.84 5	99.74 8	99.09 6	98.92 83	99.90 5	95.94 55	99.98 2	98.95 18	99.92 4	99.79 16

F-COLMAP		99.19 18	99.04 22	99.64 23	99.78 13	99.27 30	99.42 61	99.54 36	97.29 70	99.41 44	99.59 73	98.42 19	99.93 20	98.19 54	99.69 43	99.73 22

3Dnovator+	97.12 13	99.18 19	98.97 31	99.82 6	99.17 93	99.68 8	99.81 7	99.51 47	99.20 4	98.72 96	99.89 7	95.68 63	99.97 5	98.86 25	99.86 22	99.81 10

MVSFormer		99.17 20	99.12 16	99.29 53	99.51 50	98.94 47	99.88 1	99.46 66	97.55 56	99.80 3	99.65 61	97.39 35	99.28 109	99.03 11	99.85 24	99.65 38

sss		99.17 20	99.05 20	99.53 34	99.62 39	98.97 45	99.36 69	99.62 18	97.83 43	99.67 13	99.65 61	97.37 38	99.95 13	99.19 6	99.19 54	99.68 32

DP-MVS		99.16 22	98.95 34	99.78 11	99.77 15	99.53 16	99.41 63	99.50 51	97.03 83	99.04 76	99.88 10	97.39 35	99.92 22	98.66 35	99.90 12	99.87 1

CNLPA		99.14 23	98.99 29	99.59 27	99.58 45	99.41 24	99.16 95	99.44 73	98.45 19	99.19 64	99.49 87	98.08 24	99.89 37	97.73 74	99.75 39	99.48 60

MG-MVS		99.13 24	99.02 27	99.45 44	99.57 47	98.63 60	99.07 102	99.34 87	98.99 9	99.61 23	99.82 30	97.98 27	99.87 41	97.00 96	99.80 35	99.85 3

EPP-MVSNet		99.13 24	98.99 29	99.53 34	99.65 36	99.06 39	99.81 7	99.33 88	97.43 64	99.60 24	99.88 10	97.14 41	99.84 48	99.13 9	98.94 60	99.69 29

jason		99.13 24	99.03 24	99.45 44	99.46 56	98.87 51	99.12 98	99.26 93	98.03 37	99.79 5	99.65 61	97.02 44	99.85 46	99.02 13	99.90 12	99.65 38
jason: jason.
lupinMVS		99.13 24	99.01 28	99.46 43	99.51 50	98.94 47	99.05 105	99.16 100	97.86 41	99.80 3	99.56 81	97.39 35	99.86 43	98.94 19	99.85 24	99.58 48

Vis-MVSNet		99.12 28	98.97 31	99.56 31	99.78 13	99.10 35	99.68 20	99.66 15	98.49 18	99.86 1	99.87 15	94.77 73	99.84 48	99.19 6	99.41 51	99.74 20
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
TAMVS		99.12 28	99.08 18	99.24 55	99.46 56	98.55 61	99.51 43	99.46 66	98.09 30	99.45 39	99.82 30	98.34 21	99.51 99	98.70 32	98.93 61	99.67 35

DP-MVS Recon		99.12 28	98.95 34	99.65 20	99.74 21	99.70 7	99.27 84	99.57 24	96.40 107	99.42 42	99.68 58	98.75 9	99.80 57	97.98 65	99.72 40	99.44 65

CPTT-MVS		99.11 31	98.90 36	99.74 13	99.80 12	99.46 22	99.59 27	99.49 54	97.03 83	99.63 19	99.69 57	97.27 39	99.96 8	97.82 69	99.84 29	99.81 10

CDS-MVSNet		99.09 32	99.03 24	99.25 54	99.42 60	98.73 57	99.45 48	99.46 66	98.11 28	99.46 38	99.77 44	98.01 26	99.37 103	98.70 32	98.92 64	99.66 36
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
PVSNet_Blended		99.08 33	98.97 31	99.42 46	99.76 17	98.79 55	98.78 121	99.91 2	96.74 91	99.67 13	99.49 87	97.53 33	99.88 39	98.98 16	99.85 24	99.60 45

OMC-MVS		99.08 33	99.04 22	99.20 57	99.67 34	98.22 68	99.28 82	99.52 42	98.07 32	99.66 16	99.81 34	97.79 30	99.78 61	97.79 71	99.81 34	99.60 45

WTY-MVS		99.06 35	98.88 37	99.61 25	99.62 39	99.16 34	99.37 67	99.56 26	98.04 36	99.53 31	99.62 71	96.84 46	99.94 14	98.85 26	98.49 82	99.72 26

IS-MVSNet		99.05 36	98.87 38	99.57 29	99.73 23	99.32 26	99.75 14	99.20 97	98.02 38	99.56 27	99.86 18	96.54 49	99.67 85	98.09 60	99.13 56	99.73 22

API-MVS		99.04 37	99.03 24	99.06 60	99.40 65	99.31 28	99.55 39	99.56 26	98.54 15	99.33 51	99.39 97	98.76 8	99.78 61	96.98 98	99.78 36	98.07 131

PAPM_NR		99.04 37	98.84 40	99.66 17	99.74 21	99.44 23	99.39 66	99.38 82	97.70 49	99.28 54	99.28 109	98.34 21	99.85 46	96.96 99	99.45 50	99.69 29

PLC	97.94 4	99.02 39	98.85 39	99.53 34	99.66 35	99.01 43	99.24 90	99.52 42	96.85 88	99.27 56	99.48 89	98.25 23	99.91 28	97.76 72	99.62 47	99.65 38
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
AdaColmap		99.01 40	98.80 42	99.66 17	99.56 48	99.54 15	99.18 94	99.70 11	98.18 26	99.35 50	99.63 66	96.32 51	99.90 31	97.48 81	99.77 37	99.55 50

1112_ss		98.98 41	98.77 45	99.59 27	99.68 33	99.02 41	99.25 88	99.48 56	97.23 74	99.13 67	99.58 76	96.93 45	99.90 31	98.87 23	98.78 72	99.84 5

MSDG		98.98 41	98.80 42	99.53 34	99.76 17	99.19 32	98.75 124	99.55 30	97.25 72	99.47 37	99.77 44	97.82 29	99.87 41	96.93 101	99.90 12	99.54 52

114514_t		98.93 43	98.67 51	99.72 15	99.85 6	99.53 16	99.62 21	99.59 21	92.65 128	99.71 10	99.78 42	98.06 25	99.90 31	98.84 27	99.91 6	99.74 20

Test_1112_low_res		98.89 44	98.66 52	99.57 29	99.69 32	98.95 46	99.03 108	99.47 65	96.98 85	99.15 66	99.23 111	96.77 47	99.89 37	98.83 28	98.78 72	99.86 2

UGNet		98.87 45	98.69 49	99.40 47	99.22 85	98.72 58	99.44 52	99.68 14	99.24 3	99.18 65	99.42 92	92.74 95	99.96 8	99.34 4	99.94 3	99.53 55
Wanjuan Su, Qingshan Xu, Wenbing Tao: Uncertainty-guided Multi-view Stereo Network for Depth Estimation. IEEE Transactions on Circuits and Systems for Video Technology, 2022
Vis-MVSNet (Re-imp)		98.87 45	98.72 47	99.31 51	99.71 27	98.88 50	99.80 10	99.44 73	97.91 40	99.36 49	99.78 42	95.49 64	99.43 102	97.91 66	99.11 57	99.62 44

COLMAP_ROB	97.56 6	98.86 47	98.75 46	99.17 58	99.88 4	98.53 62	99.34 73	99.59 21	97.55 56	98.70 100	99.89 7	95.83 58	99.90 31	98.10 59	99.90 12	99.08 84
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
MAR-MVS		98.86 47	98.63 54	99.54 32	99.37 68	99.66 11	99.45 48	99.54 36	96.61 96	99.01 80	99.40 95	97.09 42	99.86 43	97.68 77	99.53 49	99.10 81
Zhenyu Xu, Yiguang Liu, Xuelei Shi, Ying Wang, Yunan Zheng: MARMVS: Matching Ambiguity Reduced Multiple View Stereo for Efficient Large Scale Scene Reconstruction. CVPR 2020
EPNet		98.86 47	98.71 48	99.30 52	97.20 140	98.18 69	99.62 21	98.91 118	99.28 2	98.63 106	99.81 34	95.96 54	99.99 1	99.24 5	99.72 40	99.73 22

ab-mvs		98.86 47	98.63 54	99.54 32	99.64 37	99.19 32	99.44 52	99.54 36	97.77 46	99.30 52	99.81 34	94.20 81	99.93 20	99.17 8	98.82 70	99.49 59

PVSNet_BlendedMVS		98.86 47	98.80 42	99.03 61	99.76 17	98.79 55	99.28 82	99.91 2	97.42 65	99.67 13	99.37 101	97.53 33	99.88 39	98.98 16	97.29 112	98.42 125

HY-MVS	97.30 7	98.85 52	98.64 53	99.47 41	99.42 60	99.08 37	99.62 21	99.36 85	97.39 67	99.28 54	99.68 58	96.44 50	99.92 22	98.37 48	98.22 84	99.40 68

PVSNet	96.02 17	98.85 52	98.84 40	98.89 66	99.73 23	97.28 79	98.32 133	99.60 19	97.86 41	99.50 34	99.57 79	96.75 48	99.86 43	98.56 38	99.70 42	99.54 52

PatchMatch-RL		98.84 54	98.62 56	99.52 38	99.71 27	99.28 29	99.06 103	99.77 7	97.74 47	99.50 34	99.53 84	95.41 65	99.84 48	97.17 89	99.64 45	99.44 65

DeepPCF-MVS	98.18 3	98.81 55	99.37 5	97.12 127	99.60 43	91.75 136	98.61 128	99.44 73	99.35 1	99.83 2	99.85 21	98.70 13	99.81 56	99.02 13	99.91 6	99.81 10

PMMVS		98.80 56	98.62 56	99.34 49	99.27 81	98.70 59	98.76 123	99.31 91	97.34 69	99.21 61	99.07 119	97.20 40	99.82 53	98.56 38	98.87 67	99.52 57

XVG-OURS		98.73 57	98.68 50	98.88 68	99.70 30	97.73 74	98.92 117	99.55 30	98.52 17	99.45 39	99.84 24	95.27 67	99.91 28	98.08 63	98.84 69	99.00 90

XVG-OURS-SEG-HR		98.69 58	98.62 56	98.89 66	99.71 27	97.74 73	99.12 98	99.54 36	98.44 21	99.42 42	99.71 52	94.20 81	99.92 22	98.54 42	98.90 66	99.00 90

QAPM		98.67 59	98.30 69	99.80 9	99.20 87	99.67 9	99.77 11	99.72 9	94.74 115	98.73 95	99.90 5	95.78 60	99.98 2	96.96 99	99.88 18	99.76 19

test_djsdf		98.67 59	98.57 60	98.98 63	98.70 125	98.91 49	99.88 1	99.46 66	97.55 56	99.22 60	99.88 10	95.73 62	99.28 109	99.03 11	97.62 96	98.75 96

PAPR		98.63 61	98.34 65	99.51 40	99.40 65	99.03 40	98.80 120	99.36 85	96.33 108	99.00 81	99.12 117	98.46 17	99.84 48	95.23 120	99.37 52	99.66 36

MVSTER		98.49 62	98.32 67	99.00 62	99.35 71	99.02 41	99.54 40	99.38 82	97.41 66	99.20 63	99.73 49	93.86 86	99.36 105	98.87 23	97.56 98	98.62 109

OpenMVS	96.50 16	98.47 63	98.12 72	99.52 38	99.04 110	99.53 16	99.82 6	99.72 9	94.56 117	98.08 120	99.88 10	94.73 74	99.98 2	97.47 82	99.76 38	99.06 86

IterMVS-LS		98.46 64	98.42 62	98.58 78	99.59 44	98.00 70	99.37 67	99.43 78	96.94 87	99.07 71	99.59 73	97.87 28	99.03 119	98.32 51	95.62 134	98.71 97
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
BH-untuned		98.42 65	98.36 63	98.59 77	99.49 55	96.70 91	99.27 84	99.13 101	97.24 73	98.80 92	99.38 98	95.75 61	99.74 66	97.07 94	99.16 55	99.33 71

BH-RMVSNet		98.41 66	98.08 75	99.40 47	99.41 62	98.83 54	99.30 78	98.77 122	97.70 49	98.94 82	99.65 61	92.91 93	99.74 66	96.52 104	99.55 48	99.64 42

XXY-MVS		98.38 67	98.09 74	99.24 55	99.26 82	99.32 26	99.56 35	99.55 30	97.45 63	98.71 97	99.83 26	93.23 89	99.63 92	98.88 21	96.32 126	98.76 95

ACMM	97.58 5	98.37 68	98.34 65	98.48 83	99.41 62	97.10 83	99.56 35	99.45 71	98.53 16	99.04 76	99.85 21	93.00 90	99.71 79	98.74 30	97.45 104	98.64 107
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
tpmrst		98.33 69	98.48 61	97.90 108	99.16 95	94.78 120	99.31 77	99.11 102	97.27 71	99.45 39	99.59 73	95.33 66	99.84 48	98.48 44	98.61 75	99.09 83

HyFIR		98.32 70	98.20 70	98.70 76	99.37 68	97.58 75	98.78 121	99.59 21	96.59 98	99.03 79	97.85 136	97.08 43	98.80 132	95.59 117	99.91 6	99.72 26

PatchmatchNet		98.31 71	98.36 63	98.19 97	99.16 95	95.32 112	99.27 84	98.92 115	97.37 68	99.37 48	99.58 76	94.90 71	99.70 83	97.43 85	99.21 53	99.54 52
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
LPG-MVS_test		98.22 72	98.13 71	98.49 81	99.33 73	97.05 84	99.58 29	99.55 30	97.46 61	99.24 58	99.83 26	92.58 97	99.72 74	98.09 60	97.51 99	98.68 99

RPSCF		98.22 72	98.62 56	96.99 128	99.82 9	91.58 137	99.72 15	99.44 73	96.61 96	99.66 16	99.89 7	95.92 56	99.82 53	97.46 83	99.10 58	99.57 49

CR-MVSNet		98.17 74	97.93 81	98.87 69	99.18 90	98.49 63	99.22 92	99.33 88	96.96 86	99.56 27	99.38 98	94.33 78	99.00 122	94.83 124	98.58 78	99.14 79

CLD-MVS		98.16 75	98.10 73	98.33 89	99.29 78	96.82 89	98.75 124	99.44 73	97.83 43	99.13 67	99.55 82	92.92 92	99.67 85	98.32 51	97.69 95	98.48 123
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
ACMH	97.28 8	98.10 76	97.99 78	98.44 86	99.41 62	96.96 87	99.60 26	99.56 26	98.09 30	98.15 118	99.91 4	90.87 112	99.70 83	98.88 21	97.45 104	98.67 102
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
ACMP	97.20 11	98.06 77	97.94 80	98.45 85	99.37 68	97.01 86	99.44 52	99.49 54	97.54 59	98.45 109	99.79 40	91.95 103	99.72 74	97.91 66	97.49 103	98.62 109
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
EPNet_dtu		98.03 78	97.96 79	98.23 95	98.27 132	95.54 108	99.23 91	98.75 123	99.02 8	97.82 123	99.71 52	96.11 53	99.48 100	93.04 127	99.65 44	99.69 29

FMVSNet3		98.03 78	97.76 88	98.84 71	99.39 67	98.98 44	99.40 65	99.38 82	96.67 94	99.07 71	99.28 109	92.93 91	98.98 125	97.10 91	96.65 119	98.56 119

LTVRE_ROB	97.16 12	98.02 80	97.90 82	98.40 87	99.23 84	96.80 90	99.70 18	99.60 19	97.12 79	98.18 117	99.70 54	91.73 105	99.72 74	98.39 47	97.45 104	98.68 99
Andreas Kuhn, Heiko Hirschmüller, Daniel Scharstein, Helmut Mayer: A TV Prior for High-Quality Scalable Multi-View Stereo Reconstruction. International Journal of Computer Vision 2016
PatchFormer-LS_test		98.01 81	98.05 77	97.87 109	99.15 97	94.76 121	99.42 61	98.93 113	97.12 79	98.84 89	98.59 132	93.74 88	99.80 57	98.55 40	98.17 85	99.06 86

BH-w/o		98.00 82	97.89 83	98.32 90	99.35 71	96.20 101	99.01 111	98.90 119	96.42 106	98.38 111	99.00 122	95.26 68	99.72 74	96.06 109	98.61 75	99.03 88

PT_06_test		97.99 83	98.08 75	97.75 114	98.82 120	95.43 111	99.57 32	99.48 56	96.83 90	98.87 86	98.79 126	95.88 57	99.15 114	98.17 55	96.29 127	98.66 105

ACMH+	97.24 10	97.92 84	97.78 85	98.32 90	99.46 56	96.68 93	99.56 35	99.54 36	98.41 22	97.79 125	99.87 15	90.18 119	99.66 87	98.05 64	97.18 115	98.62 109

OurMVSNet-221017-0		97.88 85	97.77 86	98.19 97	98.71 124	96.53 94	99.88 1	99.00 109	97.79 45	98.78 93	99.94 2	91.68 106	99.35 106	97.21 87	96.99 117	98.69 98

XVG-ACMP-BASELINE		97.83 86	97.71 90	98.20 96	99.11 101	96.33 97	99.41 63	99.52 42	98.06 35	99.05 75	99.50 86	89.64 122	99.73 72	97.73 74	97.38 110	98.53 120

IterMVS		97.83 86	97.77 86	98.02 102	99.58 45	96.27 99	99.02 109	99.48 56	97.22 75	98.71 97	99.70 54	92.75 94	99.13 115	97.46 83	96.00 130	98.67 102
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
EPMVS		97.82 88	97.65 91	98.35 88	98.88 118	95.98 103	99.49 45	94.71 144	97.57 55	99.26 57	99.48 89	92.46 101	99.71 79	97.87 68	99.08 59	99.35 69

TR-MVS		97.76 89	97.41 99	98.82 72	99.06 108	97.87 71	98.87 119	98.56 131	96.63 95	98.68 101	99.22 112	92.49 99	99.65 88	95.40 119	97.79 93	98.95 93

dp		97.75 90	97.80 84	97.59 119	99.10 104	93.71 129	99.32 74	98.88 120	96.48 104	99.08 70	99.55 82	92.67 96	99.82 53	96.52 104	98.58 78	99.24 76

Patchmtry		97.75 90	97.40 100	98.81 73	99.10 104	98.87 51	99.11 101	99.33 88	94.83 114	98.81 91	99.38 98	94.33 78	99.02 120	96.10 108	95.57 135	98.53 120

TAPA-MVS	97.07 15	97.74 92	97.34 105	98.94 64	99.70 30	97.53 76	99.25 88	99.51 47	91.90 130	99.30 52	99.63 66	98.78 6	99.64 89	88.09 139	99.87 20	99.65 38
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MIMVSNet		97.73 93	97.45 97	98.57 79	99.45 59	97.50 77	99.02 109	98.98 110	96.11 111	99.41 44	99.14 115	90.28 115	98.74 133	95.74 113	98.93 61	99.47 62

FMVSNet2		97.72 94	97.36 102	98.80 74	99.51 50	98.84 53	99.45 48	99.42 79	96.49 101	98.86 88	99.29 108	90.26 116	98.98 125	96.44 106	96.56 122	98.58 118

CostFormer		97.72 94	97.73 89	97.71 117	99.15 97	94.02 125	99.54 40	99.02 108	94.67 116	99.04 76	99.35 102	92.35 102	99.77 63	98.50 43	97.94 91	99.34 70

cascas		97.69 96	97.43 98	98.48 83	98.60 129	97.30 78	98.18 136	99.39 80	92.96 126	98.41 110	98.78 129	93.77 87	99.27 111	98.16 56	98.61 75	98.86 94

GBi-Net		97.68 97	97.48 94	98.29 92	99.51 50	97.26 80	99.43 56	99.48 56	96.49 101	99.07 71	99.32 105	90.26 116	98.98 125	97.10 91	96.65 119	98.62 109

test1		97.68 97	97.48 94	98.29 92	99.51 50	97.26 80	99.43 56	99.48 56	96.49 101	99.07 71	99.32 105	90.26 116	98.98 125	97.10 91	96.65 119	98.62 109

tpm		97.67 99	97.55 92	98.03 100	99.02 114	95.01 117	99.43 56	98.54 132	96.44 105	99.12 69	99.34 103	91.83 104	99.60 95	97.75 73	96.46 124	99.48 60

PCF-MVS	97.08 14	97.66 100	97.06 111	99.47 41	99.61 41	99.09 36	98.04 137	99.25 95	91.24 133	98.51 108	99.70 54	94.55 76	99.91 28	92.76 128	99.85 24	99.42 67
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
Test4		97.61 101	97.50 93	97.93 106	99.03 111	94.86 118	99.72 15	98.63 129	96.59 98	98.36 112	99.57 79	89.32 123	99.74 66	98.14 57	98.16 86	99.20 77

PAPM		97.59 102	97.09 110	99.07 59	99.06 108	98.26 67	98.30 134	99.10 103	94.88 113	98.08 120	99.34 103	96.27 52	99.64 89	89.87 135	98.92 64	99.31 72

PatchFormer_test		97.53 103	97.40 100	97.93 106	99.03 111	94.86 118	99.57 32	98.63 129	96.59 98	98.36 112	98.79 126	89.32 123	99.74 66	98.14 57	98.16 86	99.20 77

LF4IMVS		97.52 104	97.46 96	97.70 118	98.98 116	95.55 107	99.29 81	98.82 121	98.07 32	98.66 102	99.64 65	89.97 120	99.61 94	97.01 95	96.68 118	97.94 134

JIA-IIPM		97.50 105	97.02 112	98.93 65	98.73 121	97.80 72	99.30 78	98.97 111	91.73 131	98.91 84	94.86 141	95.10 70	99.71 79	97.58 79	97.98 90	99.28 74

SixPastTwentyTwo		97.50 105	97.33 107	98.03 100	98.65 127	96.23 100	99.77 11	98.68 128	97.14 77	97.90 122	99.93 3	90.45 114	99.18 113	97.00 96	96.43 125	98.67 102

tpm2		97.44 107	97.34 105	97.74 116	99.15 97	94.36 123	99.45 48	98.94 112	93.45 125	98.90 85	99.44 91	91.35 108	99.59 96	97.31 86	98.07 88	99.29 73

tpm cat1		97.39 108	97.36 102	97.50 122	99.17 93	93.73 127	99.43 56	99.31 91	91.27 132	98.71 97	99.08 118	94.31 80	99.77 63	96.41 107	98.50 81	99.00 90

tpmp4_e23		97.34 109	97.29 108	97.52 120	99.25 83	93.73 127	99.58 29	99.19 99	94.00 120	98.20 116	99.41 93	90.74 113	99.74 66	97.13 90	98.07 88	99.07 85

USDC		97.34 109	97.20 109	97.75 114	99.07 106	95.20 114	98.51 131	99.04 107	97.99 39	98.31 115	99.86 18	89.02 125	99.55 98	95.67 115	97.36 111	98.49 122

DSMNet-mixed		97.25 111	97.35 104	96.95 129	97.84 134	93.61 131	99.57 32	96.63 141	96.13 110	98.87 86	98.61 131	94.59 75	97.70 138	95.08 122	98.86 68	99.55 50

TinyColmap		97.12 112	96.89 113	97.83 111	99.07 106	95.52 109	98.57 129	98.74 126	97.58 54	97.81 124	99.79 40	88.16 132	99.56 97	95.10 121	97.21 113	98.39 127

K. test v3		97.10 113	96.79 115	98.01 103	98.72 123	96.33 97	99.87 4	97.05 140	97.59 52	96.16 136	99.80 37	88.71 126	99.04 117	96.69 103	96.55 123	98.65 106

LP		97.04 114	96.80 114	97.77 113	98.90 117	95.23 113	98.97 114	99.06 106	94.02 119	98.09 119	99.41 93	93.88 84	98.82 131	90.46 132	98.42 83	99.26 75

PatchT		97.03 115	96.44 116	98.79 75	98.99 115	98.34 65	99.16 95	99.07 105	92.13 129	99.52 33	97.31 138	94.54 77	98.98 125	88.54 137	98.73 74	99.03 88

FMVSNet1		96.84 116	96.36 117	98.29 92	99.32 75	97.26 80	99.43 56	99.48 56	95.11 112	98.55 107	99.32 105	83.95 138	98.98 125	95.81 112	96.26 128	98.62 109

test_0402		96.64 117	96.24 118	97.85 110	98.85 119	96.43 95	99.44 52	99.26 93	93.52 124	96.98 132	99.52 85	88.52 130	99.20 112	92.58 130	97.50 101	97.93 135

RPMNet		96.61 118	95.85 123	98.87 69	99.18 90	98.49 63	99.22 92	99.08 104	88.72 139	99.56 27	97.38 137	94.08 83	99.00 122	86.87 141	98.58 78	99.14 79

UnsupCasMVSNet_eth		96.44 119	96.12 120	97.40 124	98.65 127	95.65 105	99.36 69	99.51 47	97.13 78	96.04 137	98.99 123	88.40 131	98.17 136	96.71 102	90.27 140	98.40 126

FMVSNet5		96.43 120	96.19 119	97.15 126	99.11 101	95.89 104	99.32 74	99.52 42	94.47 118	98.34 114	99.07 119	87.54 133	97.07 139	92.61 129	95.72 132	98.47 124

new_pmnet		96.38 121	96.03 121	97.41 123	98.13 133	95.16 116	99.05 105	99.20 97	93.94 121	97.39 129	98.79 126	91.61 107	99.04 117	90.43 133	95.77 131	98.05 132

IB-MVS	95.67 18	96.22 122	95.44 127	98.57 79	99.21 86	96.70 91	98.65 127	97.74 139	96.71 93	97.27 130	98.54 133	86.03 134	99.92 22	98.47 46	86.30 142	99.10 81
Christian Sormann, Mattia Rossi, Andreas Kuhn and Friedrich Fraundorfer: IB-MVS: An Iterative Algorithm for Deep Multi-View Stereo based on Binary Decisions. BMVC 2021
PVSNet_0	94.43 19	96.09 123	95.47 126	97.94 105	99.31 76	94.34 124	97.81 138	99.70 11	97.12 79	97.46 128	98.75 130	89.71 121	99.79 60	97.69 76	81.69 144	99.68 32

EG-PatchMatch MVS		95.97 124	95.69 125	96.81 131	97.78 135	92.79 134	99.16 95	98.93 113	96.16 109	94.08 140	99.22 112	82.72 139	99.47 101	95.67 115	97.50 101	98.17 129

MVS-HIRNet		95.75 125	95.16 128	97.51 121	99.30 77	93.69 130	98.88 118	95.78 142	85.09 142	98.78 93	92.65 143	91.29 109	99.37 103	94.85 123	99.85 24	99.46 63

testpf		95.66 126	96.02 122	94.58 134	98.35 131	92.32 135	97.25 140	97.91 137	92.83 127	97.03 131	98.99 123	88.69 127	98.61 134	95.72 114	97.40 108	92.80 140

MIMVSNet1		95.51 127	95.04 129	96.92 130	97.38 136	95.60 106	99.52 42	99.50 51	93.65 123	96.97 133	99.17 114	85.28 136	96.56 141	88.36 138	95.55 136	98.60 116

TDRefinement		95.42 128	94.57 131	97.97 104	89.83 141	96.11 102	99.48 46	98.75 123	96.74 91	96.68 134	99.88 10	88.65 129	99.71 79	98.37 48	82.74 143	98.09 130

N_pmnet		94.95 129	95.83 124	92.31 137	98.47 130	79.33 142	99.12 98	92.81 148	93.87 122	97.68 127	99.13 116	93.87 85	99.01 121	91.38 131	96.19 129	98.59 117

OpenMVS_ROB	92.34 20	94.38 130	93.70 132	96.41 133	97.38 136	93.17 133	99.06 103	98.75 123	86.58 140	94.84 139	98.26 135	81.53 140	99.32 107	89.01 136	97.87 92	96.76 137

CMPMVS	69.68 23	94.13 131	94.90 130	91.84 138	97.24 139	80.01 141	98.52 130	99.48 56	89.01 137	91.99 143	99.67 60	85.67 135	99.13 115	95.44 118	97.03 116	96.39 138
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
HyFIR lowres test		93.82 132	92.65 133	97.33 125	99.03 111	93.48 132	98.72 126	97.91 137	90.29 135	97.72 126	98.28 134	62.59 143	98.99 124	90.37 134	98.93 61	99.53 55

UnsupCasMVSNet_bld		93.53 133	92.51 134	96.58 132	97.38 136	93.82 126	98.24 135	99.48 56	91.10 134	93.10 141	96.66 139	74.89 141	98.37 135	94.03 125	87.71 141	97.56 136

Gipuma		90.99 134	90.15 135	93.51 135	98.73 121	90.12 138	93.98 144	99.45 71	79.32 143	92.28 142	94.91 140	69.61 142	97.98 137	87.42 140	95.67 133	92.45 142
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
FPMVS		84.93 135	85.65 136	82.75 144	86.77 143	63.39 147	98.35 132	98.92 115	74.11 145	83.39 144	98.98 125	50.85 144	92.40 143	84.54 142	94.97 137	92.46 141

no-one		83.04 136	80.12 137	91.79 139	89.44 142	85.65 139	99.32 74	98.32 133	89.06 136	79.79 145	89.16 146	44.86 145	96.67 140	84.33 143	46.78 149	93.05 139

PNet_i23d		79.43 137	77.68 138	84.67 141	86.18 144	71.69 146	96.50 142	93.68 145	75.17 144	71.33 146	91.18 144	32.18 147	90.62 144	78.57 145	74.34 145	91.71 144

ANet_high		77.30 138	74.86 140	84.62 142	75.88 148	77.61 143	97.63 139	93.15 147	88.81 138	64.27 148	89.29 145	36.51 146	83.93 147	75.89 146	52.31 148	92.33 143

MVE	76.82 21	76.91 139	74.31 141	84.70 140	85.38 146	76.05 144	96.88 141	93.17 146	67.39 147	71.28 147	89.01 147	21.66 151	87.69 146	71.74 148	72.29 146	90.35 145
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
PMVS	70.75 22	75.98 140	74.97 139	79.01 145	70.98 149	55.18 148	93.37 145	98.21 134	65.08 148	61.78 150	93.83 142	21.74 150	92.53 142	78.59 144	91.12 139	89.34 146
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
wuykxyi23d		74.42 141	71.19 142	84.14 143	76.16 147	74.29 145	96.00 143	92.57 149	69.57 146	63.84 149	87.49 148	21.98 148	88.86 145	75.56 147	57.50 147	89.26 147

wuyk23d		40.18 142	41.29 143	36.84 146	86.18 144	49.12 149	79.73 146	22.81 150	27.64 149	25.46 151	28.45 149	21.98 148	48.89 148	55.80 149	23.56 150	12.51 148

ab-mvs-re		8.30 143	11.06 144	0.00 147	0.00 150	0.00 150	0.00 147	0.00 151	0.00 150	0.00 152	99.58 76	0.00 152	0.00 149	0.00 150	0.00 151	0.00 149

DeepMVS_CX				93.34 136	99.29 78	82.27 140		99.22 96	85.15 141	96.33 135	99.05 121	90.97 111	99.73 72	93.57 126	97.77 94	98.01 133

HQP4-MVS										98.66 102			99.64 89			98.64 107

lessismore_v0				97.79 112	98.69 126	95.44 110		94.75 143		95.71 138	99.87 15	88.69 127	99.32 107	95.89 110	94.93 138	98.62 109

LGP-MVS_train				98.49 81	99.33 73	97.05 84		99.55 30	97.46 61	99.24 58	99.83 26	92.58 97	99.72 74	98.09 60	97.51 99	98.68 99

door								97.92 136

HQP5-MVS						96.83 88

HQP-NCC					99.19 88		98.98 112		98.24 24	98.66 102

ACMP_Plane					99.19 88		98.98 112		98.24 24	98.66 102

BP-MVS														97.19 88

HQP3-MVS								99.39 80							97.58 97

ITE_SJBPF				98.08 99	99.29 78	96.37 96		98.92 115	98.34 23	98.83 90	99.75 46	91.09 110	99.62 93	95.82 111	97.40 108	98.25 128

HQP2-MVS												92.47 100

NP-MVS											99.40 95

MDTV_nov1_ep13_2view						95.18 115	99.35 72		96.84 89	99.58 25		95.19 69		97.82 69		99.46 63

MDTV_nov1_ep13			98.32 67		99.11 101	94.44 122	99.27 84	98.74 126	97.51 60	99.40 47	99.62 71	94.78 72	99.76 65	97.59 78	98.81 71

ACMMP++_ref															97.19 114

ACMMP++															97.43 107

HQP-MVS												85.22 137

Test By Simon												98.75 9

door-mid								98.05 135

High-resolution multi-view benchmark