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.

TDRefinement		93.52 1	93.39 2	93.88 1	95.94 2	90.26 3	95.70 2	96.46 1	90.58 2	92.86 11	96.29 4	88.16 11	94.17 21	86.07 5	98.48 7	97.22 1

LTVRE_ROB	86.10 1	93.04 2	93.44 1	91.82 5	93.73 12	85.72 9	96.79 1	95.51 3	88.86 5	95.63 2	96.99 2	84.81 18	93.16 33	91.10 1	97.53 17	96.58 4
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
COLMAP_ROB	83.01 3	91.97 3	91.95 3	92.04 2	93.68 13	86.15 7	93.37 5	95.10 4	90.28 3	92.11 14	95.03 13	89.75 8	94.93 14	79.95 20	98.27 12	95.04 13
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
ACMMP		91.91 4	91.87 4	92.03 3	95.53 3	85.91 8	93.35 6	94.16 9	82.52 14	92.39 13	94.14 21	89.15 9	95.62 6	87.35 3	98.24 13	94.56 15
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		91.67 5	91.58 6	91.96 4	95.29 4	87.62 5	93.38 4	93.36 15	83.16 13	91.06 17	94.00 23	88.26 10	95.71 5	87.28 4	98.39 9	92.55 30

LPG-MVS_test		91.47 6	91.68 5	90.82 9	94.75 7	81.69 21	90.00 14	94.27 6	82.35 15	93.67 8	94.82 16	91.18 2	95.52 7	85.36 6	98.73 5	95.23 11

SteuartSystems-ACMMP		91.04 7	91.18 8	90.60 12	93.83 11	81.72 20	90.66 11	93.61 13	82.15 17	92.63 12	93.72 26	87.29 14	95.73 4	88.35 2	98.39 9	94.36 18
Skip Steuart: Steuart Systems R&D Blog.
ACMH+	77.89 11	90.73 8	91.50 7	88.44 22	93.00 15	76.26 37	89.65 17	95.55 2	87.72 9	93.89 6	94.94 14	91.62 1	93.44 28	78.35 24	98.76 3	95.61 8

ACMM	79.39 9	90.65 9	90.99 10	89.63 17	95.03 6	83.53 17	89.62 18	93.35 16	79.20 25	93.83 7	93.60 27	90.81 5	92.96 34	85.02 8	98.45 8	92.41 32
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
LS3D		90.60 10	90.34 13	91.38 6	89.03 51	84.23 15	93.58 3	94.68 5	90.65 1	90.33 19	93.95 24	84.50 20	95.37 9	80.87 17	95.50 37	94.53 17

ACMP	79.16 10	90.54 11	90.60 12	90.35 14	94.36 9	80.98 26	89.16 22	94.05 11	79.03 27	92.87 10	93.74 25	90.60 7	95.21 12	82.87 13	98.76 3	94.87 14
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
PMVS	80.48 6	90.08 12	90.66 11	88.34 24	96.71 1	92.97 2	90.31 12	89.57 48	88.51 7	90.11 20	95.12 12	90.98 4	88.92 69	77.55 29	97.07 23	83.13 98
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
3Dnovator+	83.92 2	89.97 13	89.66 14	90.92 8	91.27 32	81.66 24	91.25 9	94.13 10	88.89 4	88.83 31	94.26 20	77.55 42	95.86 2	84.88 9	95.87 33	95.24 10

DNCC-MVS		89.54 14	89.63 15	89.26 20	92.57 17	81.34 25	90.19 13	93.08 18	80.87 21	91.13 16	93.19 28	86.22 16	95.97 1	82.23 15	97.18 21	90.45 51

CPTT-MVS		89.39 15	88.98 19	90.63 11	95.09 5	86.95 6	92.09 7	92.30 27	79.74 23	87.50 35	92.38 40	81.42 29	93.28 31	83.07 12	97.24 20	91.67 41

ACMH	76.49 14	89.34 16	91.14 9	83.96 49	92.50 19	70.36 55	89.55 19	93.84 12	81.89 18	94.70 4	95.44 10	90.69 6	88.31 75	83.33 11	98.30 11	93.20 27
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
DeepC-MVS	82.31 4	89.15 17	89.08 17	89.37 19	93.64 14	79.07 30	88.54 26	94.20 8	73.53 48	89.71 22	94.82 16	85.09 17	95.77 3	84.17 10	98.03 14	93.26 25
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
test_0402		88.65 18	89.58 16	85.88 36	92.55 18	72.22 52	84.01 52	89.44 50	88.63 6	94.38 5	95.77 6	86.38 15	93.59 27	79.84 21	95.21 42	91.82 38

DP-MVS		88.60 19	89.01 18	87.36 30	91.30 30	77.50 33	87.55 30	92.97 20	87.95 8	89.62 25	92.87 33	84.56 19	93.89 23	77.65 28	96.62 27	90.70 47

wuykxyi23d		88.46 20	88.80 20	87.44 29	90.96 35	93.03 1	85.85 44	81.96 84	74.58 43	98.58 1	97.29 1	87.73 12	87.31 78	82.84 14	99.41 1	81.99 107

OMC-MVS		88.19 21	87.52 21	90.19 15	91.94 25	81.68 23	87.49 31	93.17 17	76.02 39	88.64 32	91.22 50	84.24 21	93.37 29	77.97 27	97.03 24	95.52 9

RPSCF		88.00 22	86.93 22	91.22 7	90.08 44	89.30 4	89.68 16	91.11 34	79.26 24	89.68 23	94.81 19	82.44 25	87.74 76	76.54 33	88.74 82	96.61 3

NCCC		87.36 23	86.87 23	88.83 21	92.32 21	78.84 31	86.58 41	91.09 35	78.77 29	84.85 52	90.89 55	80.85 30	95.29 10	81.14 16	95.32 40	92.34 33

DeepPCF-MVS	81.24 5	87.28 24	86.21 28	90.49 13	91.48 29	84.90 13	83.41 54	92.38 26	70.25 67	89.35 29	90.68 58	82.85 24	94.57 17	79.55 22	95.95 31	92.00 37

Vis-MVSNet		86.86 25	86.58 25	87.72 27	92.09 23	77.43 34	87.35 32	92.09 28	78.87 28	84.27 55	94.05 22	78.35 38	93.65 25	80.54 19	91.58 69	92.08 36
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
IS-MVSNet		86.66 26	86.82 24	86.17 35	92.05 24	66.87 65	91.21 10	88.64 54	86.30 10	89.60 27	92.59 36	69.22 68	94.91 15	73.89 38	97.89 16	96.72 2

PHI-MVS		86.38 27	85.81 29	88.08 25	88.44 58	77.34 35	89.35 21	93.05 19	73.15 53	84.76 53	87.70 80	78.87 36	94.18 20	80.67 18	96.29 28	92.73 29

CSCG		86.26 28	86.47 26	85.60 39	90.87 37	74.26 43	87.98 29	91.85 31	80.35 22	89.54 28	88.01 77	79.09 34	92.13 44	75.51 35	95.06 46	90.41 52

DeepC-MVS_fast	80.27 8	86.23 29	85.65 31	87.96 26	91.30 30	76.92 36	87.19 33	91.99 29	70.56 65	84.96 49	90.69 57	80.01 33	95.14 13	78.37 23	95.78 34	91.82 38
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
TAPA-MVS	77.73 12	85.71 30	84.83 35	88.37 23	88.78 54	79.72 29	87.15 35	93.50 14	69.17 68	85.80 44	89.56 65	80.76 31	92.13 44	73.21 43	95.51 36	93.25 26
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
EPP-MVSNet		85.47 31	85.04 33	86.77 31	91.52 28	69.37 57	91.63 8	87.98 61	81.51 20	87.05 36	91.83 47	66.18 75	95.29 10	70.75 46	96.89 25	95.64 7

F-COLMAP		84.97 32	83.42 42	89.63 17	92.39 20	83.40 18	88.83 23	91.92 30	73.19 52	80.18 79	89.15 68	77.04 44	93.28 31	65.82 64	92.28 66	92.21 35

3Dnovator	80.37 7	84.80 33	84.71 36	85.06 42	86.36 71	74.71 42	88.77 24	90.00 45	75.65 40	84.96 49	93.17 29	74.06 56	91.19 52	78.28 25	91.09 71	89.29 59

IterMVS-LS		84.73 34	84.98 34	83.96 49	87.35 61	63.66 72	83.25 56	89.88 47	76.06 38	89.62 25	92.37 43	73.40 59	92.52 39	78.16 26	94.77 47	95.69 6
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
MVS_111021_HR		84.63 35	84.34 38	85.49 40	90.18 43	75.86 39	79.23 82	87.13 65	73.35 49	85.56 46	89.34 66	83.60 22	90.50 63	76.64 32	94.05 52	90.09 55

FMVSNet1		84.55 36	85.45 32	81.85 58	90.27 42	61.05 84	86.83 37	88.27 57	78.57 30	89.66 24	95.64 7	75.43 50	90.68 59	69.09 53	95.33 39	93.82 19

Gipuma		84.44 37	86.33 27	78.78 76	84.20 84	73.57 44	89.55 19	90.44 40	84.24 11	84.38 54	94.89 15	76.35 48	80.40 100	76.14 34	96.80 26	82.36 105
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
MVS_111021_LR		84.28 38	83.76 40	85.83 38	89.23 49	83.07 19	80.99 72	83.56 79	72.71 55	86.07 40	89.07 69	81.75 28	86.19 84	77.11 31	93.36 56	88.24 64

EG-PatchMatch MVS		84.08 39	84.11 39	83.98 48	92.22 22	72.61 48	82.20 66	87.02 67	72.63 56	88.86 30	91.02 53	78.52 37	91.11 54	73.41 42	91.09 71	88.21 65

DP-MVS Recon		84.05 40	83.22 43	86.52 32	91.73 26	75.27 40	83.23 57	92.40 25	72.04 60	82.04 63	88.33 74	77.91 40	93.95 22	66.17 60	95.12 44	90.34 53

AdaColmap		83.66 41	83.69 41	83.57 51	90.05 45	72.26 51	86.29 42	90.00 45	78.19 32	81.65 66	87.16 84	83.40 23	94.24 19	61.69 77	94.76 48	84.21 90

MIMVSNet1		83.63 42	84.59 37	80.74 67	94.06 10	62.77 78	82.72 58	84.53 78	77.57 35	90.34 18	95.92 5	76.88 46	85.83 88	61.88 75	97.42 18	93.62 22

CNLPA		83.55 43	83.10 45	84.90 43	89.34 48	83.87 16	84.54 50	88.77 52	79.09 26	83.54 58	88.66 72	74.87 54	81.73 98	66.84 59	92.29 65	89.11 60

PAPM_NR		83.23 44	83.19 44	83.33 52	90.90 36	65.98 68	88.19 28	90.78 36	78.13 33	80.87 73	87.92 78	73.49 58	92.42 40	70.07 49	88.40 83	91.60 42

CLD-MVS		83.18 45	82.64 46	84.79 44	89.05 50	67.82 63	77.93 87	92.52 23	68.33 71	85.07 48	81.54 109	82.06 27	92.96 34	69.35 52	97.91 15	93.57 23
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
ANet_high		83.17 46	85.68 30	75.65 94	81.24 92	45.26 119	79.94 77	92.91 21	83.83 12	91.33 15	96.88 3	80.25 32	85.92 86	68.89 55	95.89 32	95.76 5

114514_t		83.10 47	82.54 49	84.77 45	92.90 16	69.10 59	86.65 40	90.62 38	54.66 103	81.46 68	90.81 56	76.98 45	94.38 18	72.62 44	96.18 29	90.82 46

UGNet		82.78 48	81.64 55	86.21 34	86.20 73	76.24 38	86.86 36	85.68 75	77.07 36	73.76 98	92.82 34	69.64 66	91.82 47	69.04 54	93.69 55	90.56 49
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
LF4IMVS		82.75 49	81.93 54	85.19 41	82.08 89	80.15 28	85.53 46	88.76 53	68.01 73	85.58 45	87.75 79	71.80 65	86.85 80	74.02 37	93.87 54	88.58 63

QAPM		82.59 50	82.59 48	82.58 55	86.44 66	66.69 66	89.94 15	90.36 43	67.97 74	84.94 51	92.58 38	72.71 62	92.18 43	70.63 48	87.73 87	88.85 62

API-MVS		82.28 51	82.61 47	81.30 63	86.29 72	69.79 56	88.71 25	87.67 63	78.42 31	82.15 62	84.15 99	77.98 39	91.59 49	65.39 65	92.75 62	82.51 104

PCF-MVS	74.62 15	82.15 52	80.92 60	85.84 37	89.43 47	72.30 50	80.53 74	91.82 32	57.36 97	87.81 33	89.92 63	77.67 41	93.63 26	58.69 85	95.08 45	91.58 43
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
PLC	73.85 16	82.09 53	80.31 63	87.45 28	90.86 38	80.29 27	85.88 43	90.65 37	68.17 72	76.32 89	86.33 89	73.12 61	92.61 38	61.40 79	90.02 78	89.44 57
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
test1		82.02 54	82.07 51	81.85 58	86.38 68	61.05 84	86.83 37	88.27 57	72.43 57	86.00 41	95.64 7	63.78 80	90.68 59	65.95 61	93.34 57	93.82 19

GBi-Net		82.02 54	82.07 51	81.85 58	86.38 68	61.05 84	86.83 37	88.27 57	72.43 57	86.00 41	95.64 7	63.78 80	90.68 59	65.95 61	93.34 57	93.82 19

OpenMVS	76.72 13	81.98 56	82.00 53	81.93 57	84.42 81	68.22 61	88.50 27	89.48 49	66.92 76	81.80 65	91.86 45	72.59 63	90.16 66	71.19 45	91.25 70	87.40 73

PVSNet_Blended_VisFu		81.55 57	80.49 62	84.70 47	91.58 27	73.24 46	84.21 51	91.67 33	62.86 86	80.94 71	87.16 84	67.27 72	92.87 36	69.82 50	88.94 81	87.99 68

DELS-MVS		81.44 58	81.25 57	82.03 56	84.27 83	62.87 77	76.47 98	92.49 24	70.97 64	81.64 67	83.83 100	75.03 52	92.70 37	74.29 36	92.22 68	90.51 50

FMVSNet2		81.31 59	81.61 56	80.41 69	86.38 68	58.75 91	83.93 53	86.58 69	72.43 57	87.65 34	92.98 31	63.78 80	90.22 65	66.86 58	93.92 53	92.27 34

TinyColmap		81.25 60	82.34 50	77.99 83	85.33 75	60.68 87	82.32 63	88.33 56	71.26 63	86.97 37	92.22 44	77.10 43	86.98 79	62.37 73	95.17 43	86.31 79

BH-untuned		80.96 61	80.99 58	80.84 66	88.55 56	68.23 60	80.33 75	88.46 55	72.79 54	86.55 38	86.76 88	74.72 55	91.77 48	61.79 76	88.99 80	82.52 103

BH-RMVSNet		80.53 62	80.22 64	81.49 62	87.19 63	66.21 67	77.79 89	86.23 71	74.21 44	83.69 56	88.50 73	73.25 60	90.75 56	63.18 71	87.90 86	87.52 71

EPNet		80.37 63	78.41 69	86.23 33	76.75 113	73.28 45	87.18 34	77.45 90	76.24 37	68.14 113	88.93 70	65.41 77	93.85 24	69.47 51	96.12 30	91.55 44

MG-MVS		80.32 64	80.94 59	78.47 79	88.18 59	52.62 104	82.29 64	85.01 77	72.01 61	79.24 81	92.54 39	69.36 67	93.36 30	70.65 47	89.19 79	89.45 56

MAR-MVS		80.24 65	78.74 67	84.73 46	86.87 65	78.18 32	85.75 45	87.81 62	65.67 79	77.84 83	78.50 119	73.79 57	90.53 62	61.59 78	90.87 75	85.49 83
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
MSDG		80.06 66	79.99 65	80.25 70	83.91 85	68.04 62	77.51 92	89.19 51	77.65 34	81.94 64	83.45 103	76.37 47	86.31 83	63.31 70	86.59 92	86.41 77

ab-mvs		79.67 67	80.56 61	76.99 87	88.48 57	56.93 93	84.70 49	86.06 73	68.95 70	80.78 74	93.08 30	75.30 51	84.62 90	56.78 91	90.90 74	89.43 58

PAPR		78.84 68	78.10 70	81.07 65	85.17 76	60.22 89	82.21 65	90.57 39	62.51 87	75.32 95	84.61 97	74.99 53	92.30 42	59.48 83	88.04 85	90.68 48

FMVSNet3		78.80 69	78.55 68	79.57 75	82.89 88	56.89 94	81.76 69	85.77 74	69.04 69	86.00 41	90.44 59	51.75 104	90.09 67	65.95 61	93.34 57	91.72 40

PVSNet_BlendedMVS		78.80 69	77.84 71	81.65 61	84.43 79	63.41 73	79.49 79	90.44 40	61.70 90	75.43 93	87.07 86	69.11 69	91.44 50	60.68 81	92.24 67	90.11 54

TAMVS		78.08 71	76.36 76	83.23 53	90.62 40	72.87 47	79.08 83	80.01 87	61.72 89	81.35 70	86.92 87	63.96 79	88.78 70	50.61 105	93.01 60	88.04 67

Vis-MVSNet (Re-imp)		77.82 72	77.79 72	77.92 84	88.82 53	51.29 108	83.28 55	71.97 109	74.04 45	82.23 61	89.78 64	57.38 95	89.41 68	57.22 89	95.41 38	93.05 28

OpenMVS_ROB	70.19 17	77.77 73	77.46 73	78.71 77	84.39 82	61.15 82	81.18 71	82.52 81	62.45 88	83.34 59	87.37 83	66.20 74	88.66 73	64.69 67	85.02 97	86.32 78

CDS-MVSNet		77.32 74	75.40 83	83.06 54	89.00 52	72.48 49	77.90 88	82.17 83	60.81 91	78.94 82	83.49 102	59.30 89	88.76 71	54.64 99	92.37 64	87.93 70
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
MVSTER		77.09 75	75.70 82	81.25 64	75.27 119	61.08 83	77.49 93	85.07 76	60.78 92	86.55 38	88.68 71	43.14 120	90.25 64	73.69 39	90.67 76	92.42 31

IterMVS		76.91 76	76.34 77	78.64 78	80.91 93	64.03 71	76.30 99	79.03 88	64.88 81	83.11 60	89.16 67	59.90 87	84.46 91	68.61 57	85.15 96	87.42 72
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
TR-MVS		76.77 77	75.79 80	79.72 73	86.10 74	65.79 70	77.14 94	83.02 80	65.20 80	81.40 69	82.10 107	66.30 73	90.73 58	55.57 95	85.27 94	82.65 100

USDC		76.63 78	76.73 75	76.34 92	83.46 86	57.20 92	80.02 76	88.04 60	52.14 111	83.65 57	91.25 49	63.24 83	86.65 82	54.66 98	94.11 51	85.17 84

BH-w/o		76.57 79	76.07 79	78.10 81	86.88 64	65.92 69	77.63 90	86.33 70	65.69 78	80.89 72	79.95 115	68.97 71	90.74 57	53.01 103	85.25 95	77.62 116

Patchmtry		76.56 80	77.46 73	73.83 99	79.37 101	46.60 118	82.41 61	76.90 93	73.81 46	85.56 46	92.38 40	48.07 112	83.98 94	63.36 69	95.31 41	90.92 45

PVSNet_Blended		76.49 81	75.40 83	79.76 72	84.43 79	63.41 73	75.14 103	90.44 40	57.36 97	75.43 93	78.30 120	69.11 69	91.44 50	60.68 81	87.70 88	84.42 88

cascas		76.29 82	74.81 85	80.72 68	84.47 78	62.94 76	73.89 109	87.34 64	55.94 100	75.16 97	76.53 123	63.97 78	91.16 53	65.00 66	90.97 73	88.06 66

RPMNet		76.06 83	75.79 80	76.85 89	79.58 97	62.64 79	82.58 59	71.75 112	74.80 42	75.72 91	92.59 36	48.69 110	84.07 92	73.48 41	82.91 103	83.85 91

wuyk23d		75.13 84	79.30 66	62.63 118	75.56 116	75.18 41	80.89 73	73.10 104	75.06 41	94.76 3	95.32 11	87.73 12	52.85 127	34.16 129	97.11 22	59.85 124

CMPMVS	59.41 20	75.12 85	73.57 87	79.77 71	75.84 115	67.22 64	81.21 70	82.18 82	50.78 117	76.50 87	87.66 81	55.20 99	82.99 96	62.17 74	90.64 77	89.09 61
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
1112_ss		74.82 86	73.74 86	78.04 82	89.57 46	60.04 90	76.49 97	87.09 66	54.31 104	73.66 99	79.80 116	60.25 85	86.76 81	58.37 86	84.15 99	87.32 74

PatchMatch-RL		74.48 87	73.22 88	78.27 80	87.70 60	85.26 10	75.92 100	70.09 115	64.34 82	76.09 90	81.25 110	65.87 76	78.07 104	53.86 100	83.82 100	71.48 118

XXY-MVS		74.44 88	76.19 78	69.21 108	84.61 77	52.43 105	71.70 113	77.18 91	60.73 93	80.60 75	90.96 54	75.44 49	69.35 116	56.13 93	88.33 84	85.86 81

CR-MVSNet		74.00 89	73.04 90	76.85 89	79.58 97	62.64 79	82.58 59	76.90 93	50.50 119	75.72 91	92.38 40	48.07 112	84.07 92	68.72 56	82.91 103	83.85 91

Test_1112_low_res		73.90 90	73.08 89	76.35 91	90.35 41	55.95 96	73.40 111	86.17 72	50.70 118	73.14 100	85.94 91	58.31 92	85.90 87	56.51 92	83.22 101	87.20 75

HY-MVS	64.64 18	73.03 91	72.47 92	74.71 96	83.36 87	54.19 97	82.14 67	81.96 84	56.76 99	69.57 110	86.21 90	60.03 86	84.83 89	49.58 110	82.65 105	85.11 85

EPNet_dtu		72.87 92	71.33 97	77.49 86	77.72 108	60.55 88	82.35 62	75.79 98	66.49 77	58.39 129	81.06 111	53.68 103	85.98 85	53.55 101	92.97 61	85.95 80

FPMVS		72.29 93	72.00 94	73.14 102	88.63 55	85.00 11	74.65 106	67.39 117	71.94 62	77.80 84	87.66 81	50.48 107	75.83 109	49.95 107	79.51 113	58.58 126

FMVSNet5		72.10 94	71.69 95	73.32 100	81.57 91	53.02 101	76.77 95	78.37 89	63.31 84	76.37 88	91.85 46	36.68 127	78.98 103	47.87 116	92.45 63	87.95 69

PAPM		71.77 95	70.06 99	76.92 88	86.39 67	53.97 98	76.62 96	86.62 68	53.44 108	63.97 121	84.73 96	57.79 94	92.34 41	39.65 123	81.33 108	84.45 87

IB-MVS	62.13 19	71.64 96	68.97 101	79.66 74	80.80 96	62.26 81	73.94 108	76.90 93	63.27 85	68.63 112	76.79 122	33.83 129	91.84 46	59.28 84	87.26 89	84.88 86
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
UnsupCasMVSNet_eth		71.63 97	72.30 93	69.62 106	76.47 114	52.70 103	70.03 116	80.97 86	59.18 95	79.36 80	88.21 75	60.50 84	69.12 117	58.33 87	77.62 118	87.04 76

no-one		71.52 98	70.43 98	74.81 95	78.45 106	63.41 73	57.73 126	77.03 92	51.46 114	77.17 86	90.33 60	54.96 101	80.35 101	47.41 117	99.29 2	80.68 113

MIMVSNet		71.09 99	71.59 96	69.57 107	87.23 62	50.07 113	78.91 84	71.83 110	60.20 94	71.26 104	91.76 48	55.08 100	76.09 107	41.06 122	87.02 91	82.54 102

PatchT		70.52 100	72.76 91	63.79 117	79.38 100	33.53 128	77.63 90	65.37 120	73.61 47	71.77 103	92.79 35	44.38 119	75.65 111	64.53 68	85.37 93	82.18 106

N_pmnet		70.20 101	68.80 103	74.38 97	80.91 93	84.81 14	59.12 125	76.45 97	55.06 101	75.31 96	82.36 106	55.74 97	54.82 126	47.02 119	87.24 90	83.52 94

CostFormer		69.98 102	68.68 104	73.87 98	77.14 111	50.72 111	79.26 81	74.51 102	51.94 113	70.97 106	84.75 95	45.16 118	87.49 77	55.16 97	79.23 116	83.40 95

PatchmatchNet		69.71 103	68.83 102	72.33 104	77.66 109	53.60 99	79.29 80	69.99 116	57.66 96	72.53 101	82.93 105	46.45 114	80.08 102	60.91 80	72.09 120	83.31 96
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
tpmp4_e23		69.43 104	67.33 109	75.72 93	78.53 105	52.75 102	82.13 68	74.91 100	49.23 120	66.37 118	84.17 98	41.28 123	88.67 72	49.73 109	79.63 112	85.75 82

LP		69.42 105	68.30 106	72.77 103	71.48 126	56.84 95	73.66 110	74.84 101	63.52 83	70.95 107	83.35 104	49.55 109	77.15 105	57.13 90	70.21 123	84.33 89

JIA-IIPM		69.41 106	66.64 114	77.70 85	73.19 121	71.24 53	75.67 101	65.56 119	70.42 66	65.18 120	92.97 32	33.64 130	83.06 95	53.52 102	69.61 125	78.79 115

UnsupCasMVSNet_bld		69.21 107	69.68 100	67.82 110	79.42 99	51.15 109	67.82 118	75.79 98	54.15 105	77.47 85	85.36 94	59.26 90	70.64 115	48.46 114	79.35 115	81.66 108

tpm2		68.45 108	66.83 112	73.30 101	78.93 103	48.50 116	79.76 78	71.76 111	47.50 122	69.92 109	83.60 101	42.07 122	88.40 74	48.44 115	79.51 113	83.01 99

tpm		67.95 109	68.08 108	67.55 111	78.74 104	43.53 123	75.60 102	67.10 118	54.92 102	72.23 102	88.10 76	42.87 121	75.97 108	52.21 104	80.95 111	83.15 97

WTY-MVS		67.91 110	68.35 105	66.58 112	80.82 95	48.12 117	65.96 119	72.60 105	53.67 107	71.20 105	81.68 108	58.97 91	69.06 118	48.57 113	81.67 106	82.55 101

sss		66.92 111	67.26 110	65.90 113	77.23 110	51.10 110	64.79 120	71.72 113	52.12 112	70.13 108	80.18 113	57.96 93	65.36 122	50.21 106	81.01 110	81.25 110

tpm cat1		66.76 112	65.21 116	71.42 105	77.09 112	50.62 112	78.01 86	73.68 103	44.89 125	68.64 111	79.00 118	45.51 115	82.42 97	49.91 108	70.15 124	81.23 111

HyFIR		66.60 113	67.24 111	64.68 116	72.26 124	51.46 107	62.34 122	76.80 96	49.02 121	63.68 122	85.65 92	51.09 105	75.77 110	62.95 72	81.37 107	66.82 122

PVSNet	58.17 21	66.41 114	65.63 115	68.75 109	81.96 90	49.88 114	62.19 123	72.51 107	51.03 116	68.04 114	75.34 125	50.84 106	74.77 112	45.82 120	82.96 102	81.60 109

tpmrst		66.28 115	66.69 113	65.05 115	72.82 123	39.33 124	78.20 85	70.69 114	53.16 109	67.88 115	80.36 112	48.18 111	74.75 113	58.13 88	70.79 122	81.08 112

EPMVS		62.47 116	62.63 118	62.01 119	70.63 127	38.74 125	74.76 104	52.86 125	53.91 106	67.71 116	80.01 114	39.40 125	66.60 119	55.54 96	68.81 126	80.68 113

PMMVS		61.65 117	60.38 120	65.47 114	65.40 128	69.26 58	63.97 121	61.73 123	36.80 129	60.11 126	68.43 127	59.42 88	66.35 120	48.97 112	78.57 117	60.81 123

MVS-HIRNet		61.16 118	62.92 117	55.87 123	79.09 102	35.34 127	71.83 112	57.98 124	46.56 123	59.05 128	91.14 52	49.95 108	76.43 106	38.74 124	71.92 121	55.84 127

DSMNet-mixed		60.98 119	61.61 119	59.09 122	72.88 122	45.05 121	74.70 105	46.61 129	26.20 131	65.34 119	90.32 61	55.46 98	63.12 125	41.72 121	81.30 109	69.09 121

dp		60.70 120	60.29 121	61.92 120	72.04 125	38.67 126	70.83 114	64.08 121	51.28 115	60.75 125	77.28 121	36.59 128	71.58 114	47.41 117	62.34 129	75.52 117

PVSNet_0	51.08 22	56.10 121	54.97 123	59.48 121	75.12 120	53.28 100	55.16 127	61.89 122	44.30 126	59.16 127	62.48 130	54.22 102	65.91 121	35.40 128	47.01 130	59.25 125

new_pmnet		55.69 122	57.66 122	49.76 127	75.47 118	30.59 129	59.56 124	51.45 127	43.62 127	62.49 123	75.48 124	40.96 124	49.15 128	37.39 125	72.52 119	69.55 120

PNet_i23d		52.13 123	51.24 125	54.79 125	75.56 116	45.26 119	54.54 128	52.55 126	66.95 75	57.19 130	65.82 129	13.15 133	63.40 124	36.39 127	39.04 131	55.71 128

MVE	40.22 23	51.82 124	50.47 126	55.87 123	62.66 130	51.91 106	31.61 129	39.28 130	40.65 128	50.76 132	74.98 126	56.24 96	44.67 129	33.94 130	64.11 128	71.04 119
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
HyFIR lowres test		51.64 125	51.77 124	51.24 126	64.21 129	49.39 115	68.80 117	46.62 128	28.12 130	52.64 131	65.93 128	38.34 126	64.88 123	37.16 126	64.13 127	39.44 129

ab-mvs-re		6.65 126	8.87 127	0.00 129	0.00 132	0.00 132	0.00 130	0.00 132	0.00 133	0.00 134	79.80 116	0.00 134	0.00 131	0.00 132	0.00 133	0.00 131

door								72.57 106

NP-MVS											90.17 62

HQP4-MVS										80.56 76			94.61 16			93.56 24

ACMMP++															97.35 19

BP-MVS														77.30 30

HQP5-MVS						70.66 54

HQP-NCC					91.19 33		84.77 47		73.30 50	80.55 77

DeepMVS_CX				24.13 128	32.95 131	29.49 130		21.63 131	12.07 132	37.95 133	45.07 131	30.84 131	19.21 130	17.94 131	33.06 132	23.69 130

ITE_SJBPF				90.11 16	90.72 39	84.97 12		90.30 44	81.56 19	90.02 21	91.20 51	82.40 26	90.81 55	73.58 40	94.66 49	94.56 15

HQP-MVS												30.83 132

ACMP_Plane					91.19 33		84.77 47		73.30 50	80.55 77

Test By Simon												79.09 34

MDTV_nov1_ep13_2view						27.60 131	70.76 115		46.47 124	61.27 124		45.20 117		49.18 111		83.75 93

MDTV_nov1_ep13			68.29 107		78.03 107	43.87 122	74.12 107	72.22 108	52.17 110	67.02 117	85.54 93	45.36 116	80.85 99	55.73 94	84.42 98

ACMMP++_ref															95.74 35

LGP-MVS_train				90.82 9	94.75 7	81.69 21		94.27 6	82.35 15	93.67 8	94.82 16	91.18 2	95.52 7	85.36 6	98.73 5	95.23 11

HQP2-MVS												72.10 64

HQP3-MVS								92.68 22							94.47 50

High-resolution multi-view benchmark