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	courty.	delive.	electro	facade	kicker	meadow	office	pipes	playgr.	relief	relief.	terrace	terrai.

MCST-MVS		93.81 17	94.06 19	93.53 18	96.79 24	96.85 21	95.95 14	91.69 17	92.20 27	87.17 33	90.83 28	93.41 7	91.96 15	94.49 26	93.50 32	97.61 1	97.12 23

xxxxxxxxxxxxxcwj		92.95 25	91.88 34	94.20 9	96.75 25	97.07 12	95.82 19	92.60 7	93.98 13	91.09 9	95.89 6	71.01 129	91.93 16	94.40 29	93.56 29	97.04 2	97.27 17

SF-MVS		94.61 8	94.96 10	94.20 9	96.75 25	97.07 12	95.82 19	92.60 7	93.98 13	91.09 9	95.89 6	92.54 13	91.93 16	94.40 29	93.56 29	97.04 2	97.27 17

DVP-MVS++		95.79 1	96.42 1	95.06 1	97.84 2	98.17 2	97.03 4	92.84 3	96.68 1	92.83 3	95.90 5	94.38 4	92.90 5	95.98 2	94.85 5	96.93 4	98.99 1

SED-MVS		95.61 2	96.36 2	94.73 3	96.84 20	98.15 3	97.08 3	92.92 2	95.64 3	91.84 5	95.98 4	95.33 1	92.83 7	96.00 1	94.94 3	96.90 5	98.45 3

DVP-MVS		95.56 3	96.26 3	94.73 3	96.93 17	98.19 1	96.62 7	92.81 5	96.15 2	91.73 6	95.01 8	95.31 2	93.41 1	95.95 3	94.77 8	96.90 5	98.46 2
Zhenlong Yuan, Jinguo Luo, Fei Shen, Zhaoxin Li, Cong Liu, Tianlu Mao, Zhaoqi Wang: DVP-MVS: Synergize Depth-Edge and Visibility Prior for Multi-View Stereo. AAAI2025
DPE-MVS		95.53 4	96.13 4	94.82 2	96.81 23	98.05 4	97.42 1	93.09 1	94.31 9	91.49 7	97.12 1	95.03 3	93.27 3	95.55 6	94.58 12	96.86 7	98.25 4
Kehua Chen, Zhenlong Yuan, Tianlu Mao, Zhaoqi Wang: Dual-Level Precision Edges Guided Multi-View Stereo with Accurate Planarization. AAAI2025
TSAR-MVS + GP.		92.71 29	93.91 21	91.30 36	91.96 73	96.00 42	93.43 42	87.94 42	92.53 23	86.27 41	93.57 16	91.94 19	91.44 26	93.29 46	92.89 47	96.78 8	97.15 22

DROMVSNet		89.96 47	90.77 44	89.01 56	90.54 91	95.15 56	91.34 62	81.43 106	85.27 63	83.08 54	82.83 48	87.22 50	90.97 31	94.79 17	93.38 37	96.73 9	96.71 34

ETV-MVS		89.22 53	89.76 51	88.60 62	91.60 77	94.61 67	89.48 83	83.46 85	85.20 65	81.58 62	82.75 49	82.59 67	88.80 42	94.57 24	93.28 40	96.68 10	95.31 59

APDe-MVS		95.23 5	95.69 6	94.70 5	97.12 11	97.81 6	97.19 2	92.83 4	95.06 6	90.98 11	96.47 2	92.77 11	93.38 2	95.34 9	94.21 16	96.68 10	98.17 5

MVS_0304		90.88 41	91.35 39	90.34 43	93.91 53	96.79 24	94.49 35	86.54 51	86.57 57	82.85 57	81.68 57	89.70 34	87.57 58	94.64 22	93.93 21	96.67 12	96.15 45

3Dnovator	85.17 5	90.48 43	89.90 49	91.16 38	94.88 44	95.74 47	93.82 38	85.36 58	89.28 46	87.81 29	74.34 96	87.40 49	88.56 46	93.07 49	93.74 26	96.53 13	95.71 51

TSAR-MVS + MP.		94.48 11	94.97 9	93.90 14	95.53 38	97.01 16	96.69 6	90.71 25	94.24 10	90.92 12	94.97 9	92.19 16	93.03 4	94.83 15	93.60 27	96.51 14	97.97 9
Zhenlong Yuan, Jiakai Cao, Zhaoqi Wang, Zhaoxin Li: TSAR-MVS: Textureless-aware Segmentation and Correlative Refinement Guided Multi-View Stereo. Pattern Recognition
CS-MVS		89.44 51	90.69 45	87.99 68	91.16 81	94.56 68	89.02 95	80.75 113	86.95 55	79.62 76	83.07 47	85.88 52	91.03 30	94.49 26	93.49 34	96.37 15	96.40 40

MSP-MVS		95.12 6	95.83 5	94.30 6	96.82 22	97.94 5	96.98 5	92.37 12	95.40 4	90.59 14	96.16 3	93.71 6	92.70 8	94.80 16	94.77 8	96.37 15	97.99 8
Zhenlong Yuan, Cong Liu, Fei Shen, Zhaoxin Li, Jingguo luo, Tianlu Mao and Zhaoqi Wang: MSP-MVS: Multi-granularity Segmentation Prior Guided Multi-View Stereo. AAAI2025
QAPM		89.49 50	89.58 53	89.38 53	94.73 46	95.94 43	92.35 50	85.00 61	85.69 62	80.03 73	76.97 81	87.81 47	87.87 53	92.18 64	92.10 54	96.33 17	96.40 40

MVS_111021_HR		90.56 42	91.29 40	89.70 50	94.71 47	95.63 48	91.81 58	86.38 52	87.53 53	81.29 64	87.96 33	85.43 55	87.69 55	93.90 36	92.93 45	96.33 17	95.69 52

test1111		84.86 91	84.21 97	85.61 86	91.75 76	95.14 57	88.63 100	84.57 64	81.88 86	71.21 114	65.66 144	68.51 141	81.19 97	93.74 41	92.68 51	96.31 19	91.86 126

CANet		91.33 39	91.46 37	91.18 37	95.01 41	96.71 25	93.77 39	87.39 47	87.72 52	87.26 32	81.77 55	89.73 33	87.32 62	94.43 28	93.86 22	96.31 19	96.02 47

APD-MVS		94.37 12	94.47 16	94.26 7	97.18 9	96.99 17	96.53 8	92.68 6	92.45 25	89.96 18	94.53 12	91.63 21	92.89 6	94.58 23	93.82 23	96.31 19	97.26 19
Yuesong Wang, Zhaojie Zeng and etc.: Adaptive Patch Deformation for Textureless-Resilient Multi-View Stereo. CVPR2023
CS-MVS-test		89.17 54	89.90 49	88.31 67	89.96 103	93.19 86	91.34 62	81.43 106	84.69 71	79.88 75	82.62 51	84.70 57	90.97 31	94.79 17	92.96 44	96.29 22	96.34 43

test2506		85.20 86	84.11 98	86.47 78	91.84 74	95.28 52	89.18 86	84.49 65	82.59 76	75.34 96	74.66 94	58.07 188	81.68 93	93.76 38	92.71 49	96.28 23	91.71 129

ECVR-MVS		85.25 85	84.47 94	86.16 81	91.84 74	95.28 52	89.18 86	84.49 65	82.59 76	73.49 105	66.12 138	69.28 137	81.68 93	93.76 38	92.71 49	96.28 23	91.58 136

SteuartSystems-ACMMP		94.06 14	94.65 12	93.38 20	96.97 16	97.36 9	96.12 10	91.78 15	92.05 29	87.34 31	94.42 13	90.87 25	91.87 20	95.47 8	94.59 11	96.21 25	97.77 11
Skip Steuart: Steuart Systems R&D Blog.
PHI-MVS		92.05 33	93.74 22	90.08 45	94.96 42	97.06 14	93.11 46	87.71 45	90.71 38	80.78 69	92.40 23	91.03 23	87.68 56	94.32 31	94.48 13	96.21 25	96.16 44

CNVR-MVS		94.37 12	94.65 12	94.04 12	97.29 7	97.11 11	96.00 11	92.43 11	93.45 17	89.85 20	90.92 26	93.04 9	92.59 10	95.77 5	94.82 6	96.11 27	97.42 16

baseline1		84.54 94	84.43 95	84.67 92	90.62 89	91.16 111	88.63 100	83.75 76	79.78 106	71.16 115	75.14 90	74.10 111	77.84 139	91.56 69	90.67 74	96.04 28	88.58 157

MSLP-MVS++		92.02 35	91.40 38	92.75 25	96.01 33	95.88 45	93.73 41	89.00 35	89.89 45	90.31 16	81.28 60	88.85 40	91.45 24	92.88 53	94.24 15	96.00 29	96.76 31

NCCC		93.69 20	93.66 23	93.72 17	97.37 6	96.66 30	95.93 17	92.50 10	93.40 20	88.35 26	87.36 36	92.33 15	92.18 13	94.89 14	94.09 18	96.00 29	96.91 27

ACMMP		92.03 34	92.16 32	91.87 35	95.88 35	96.55 32	94.47 36	89.49 34	91.71 32	85.26 43	91.52 25	84.48 59	90.21 36	92.82 54	91.63 58	95.92 31	96.42 38
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
zzz-MVS		93.80 18	93.45 26	94.20 9	97.53 4	96.43 37	95.88 18	91.12 21	94.09 12	92.74 4	87.68 34	90.77 26	92.04 14	94.74 20	93.56 29	95.91 32	96.85 28

SMA-MVS		94.70 7	95.35 7	93.93 13	97.57 3	97.57 8	95.98 12	91.91 14	94.50 7	90.35 15	93.46 18	92.72 12	91.89 19	95.89 4	95.22 1	95.88 33	98.10 6
Yufeng Yin; Xiaoyan Liu; Zichao Zhang: SMA-MVS: Segmentation-Guided Multi-Scale Anchor Deformation Patch Multi-View Stereo. IEEE Transactions on Circuits and Systems for Video Technology
ACMMPR		93.72 19	93.94 20	93.48 19	97.07 12	96.93 18	95.78 22	90.66 27	93.88 15	89.24 22	93.53 17	89.08 39	92.24 12	93.89 37	93.50 32	95.88 33	96.73 32

HFP-MVS		94.02 15	94.22 18	93.78 15	97.25 8	96.85 21	95.81 21	90.94 24	94.12 11	90.29 17	94.09 15	89.98 32	92.52 11	93.94 35	93.49 34	95.87 35	97.10 24

PVSNet_BlendedMVS		88.19 65	88.00 65	88.42 64	92.71 69	94.82 64	89.08 90	83.81 74	84.91 68	86.38 38	79.14 68	78.11 92	82.66 85	93.05 50	91.10 62	95.86 36	94.86 66

PVSNet_Blended		88.19 65	88.00 65	88.42 64	92.71 69	94.82 64	89.08 90	83.81 74	84.91 68	86.38 38	79.14 68	78.11 92	82.66 85	93.05 50	91.10 62	95.86 36	94.86 66

DeepC-MVS_fast	88.76 1	93.10 23	93.02 30	93.19 23	97.13 10	96.51 34	95.35 27	91.19 20	93.14 22	88.14 27	85.26 42	89.49 36	91.45 24	95.17 10	95.07 2	95.85 38	96.48 36
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
PGM-MVS		92.76 27	93.03 29	92.45 29	97.03 14	96.67 29	95.73 24	87.92 43	90.15 44	86.53 37	92.97 21	88.33 45	91.69 22	93.62 43	93.03 42	95.83 39	96.41 39

SD-MVS		94.53 10	95.22 8	93.73 16	95.69 37	97.03 15	95.77 23	91.95 13	94.41 8	91.35 8	94.97 9	93.34 8	91.80 21	94.72 21	93.99 20	95.82 40	98.07 7
Zhenlong Yuan, Jiakai Cao, Zhaoxin Li, Hao Jiang and Zhaoqi Wang: SD-MVS: Segmentation-driven Deformation Multi-View Stereo with Spherical Refinement and EM optimization. AAAI2024
DeepC-MVS	87.86 3	92.26 32	91.86 35	92.73 26	96.18 30	96.87 20	95.19 29	91.76 16	92.17 28	86.58 36	81.79 54	85.85 53	90.88 33	94.57 24	94.61 10	95.80 41	97.18 20
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
XVS					93.11 61	96.70 26	91.91 55			83.95 49		88.82 41				95.79 42

X-MVStestdata					93.11 61	96.70 26	91.91 55			83.95 49		88.82 41				95.79 42

DELS-MVS		89.71 48	89.68 52	89.74 48	93.75 55	96.22 39	93.76 40	85.84 54	82.53 78	85.05 45	78.96 71	84.24 60	84.25 78	94.91 13	94.91 4	95.78 44	96.02 47
Christian Sormann, Emanuele Santellani, Mattia Rossi, Andreas Kuhn, Friedrich Fraundorfer: DELS-MVS: Deep Epipolar Line Search for Multi-View Stereo. Winter Conference on Applications of Computer Vision (WACV), 2023
X-MVS		92.36 31	92.75 31	91.90 34	96.89 18	96.70 26	95.25 28	90.48 30	91.50 34	83.95 49	88.20 32	88.82 41	89.11 40	93.75 40	93.43 36	95.75 45	96.83 30

casdiffmvs		87.45 71	87.15 73	87.79 72	90.15 101	94.22 71	89.96 74	83.93 73	85.08 66	80.91 66	75.81 86	77.88 95	86.08 70	91.86 67	90.86 68	95.74 46	94.37 74

CP-MVS		93.25 22	93.26 27	93.24 22	96.84 20	96.51 34	95.52 25	90.61 28	92.37 26	88.88 23	90.91 27	89.52 35	91.91 18	93.64 42	92.78 48	95.69 47	97.09 25

MVSTER		86.03 79	86.12 81	85.93 83	88.62 114	89.93 130	89.33 85	79.91 128	81.87 87	81.35 63	81.07 61	74.91 106	80.66 105	92.13 65	90.10 84	95.68 48	92.80 103

3Dnovator+	86.06 4	91.60 37	90.86 43	92.47 28	96.00 34	96.50 36	94.70 33	87.83 44	90.49 40	89.92 19	74.68 93	89.35 37	90.66 34	94.02 33	94.14 17	95.67 49	96.85 28

OpenMVS	82.53 11	87.71 68	86.84 75	88.73 59	94.42 49	95.06 59	91.02 66	83.49 82	82.50 80	82.24 61	67.62 133	85.48 54	85.56 73	91.19 75	91.30 61	95.67 49	94.75 68

EIA-MVS		87.94 67	88.05 64	87.81 70	91.46 78	95.00 61	88.67 98	82.81 90	82.53 78	80.81 68	80.04 64	80.20 78	87.48 59	92.58 57	91.61 59	95.63 51	94.36 75

ET-MVSNet_ETH3D		84.65 92	85.58 87	83.56 111	74.99 212	92.62 100	90.29 71	80.38 116	82.16 83	73.01 110	83.41 45	71.10 128	87.05 65	87.77 125	90.17 83	95.62 52	91.82 127

CDPH-MVS		91.14 40	92.01 33	90.11 44	96.18 30	96.18 40	94.89 32	88.80 39	88.76 49	77.88 87	89.18 31	87.71 48	87.29 63	93.13 48	93.31 39	95.62 52	95.84 49

MP-MVS		93.35 21	93.59 24	93.08 24	97.39 5	96.82 23	95.38 26	90.71 25	90.82 37	88.07 28	92.83 22	90.29 30	91.32 28	94.03 32	93.19 41	95.61 54	97.16 21
Rongxuan Tan, Qing Wang, et al.: MP-MVS: Multi-Scale Windows PatchMatch and Planar Prior Multi-View Stereo.
FMVSNet3		84.44 97	84.64 93	84.21 100	84.32 163	90.13 125	89.85 76	80.37 117	81.17 92	75.50 92	69.63 118	79.69 84	79.62 125	89.72 102	90.52 78	95.59 55	91.58 136

HPM-MVS++		94.60 9	94.91 11	94.24 8	97.86 1	96.53 33	96.14 9	92.51 9	93.87 16	90.76 13	93.45 19	93.84 5	92.62 9	95.11 12	94.08 19	95.58 56	97.48 14

CLD-MVS		88.66 57	88.52 58	88.82 58	91.37 80	94.22 71	92.82 49	82.08 101	88.27 51	85.14 44	81.86 53	78.53 91	85.93 72	91.17 76	90.61 75	95.55 57	95.00 62
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
ACMMP_NAP		93.94 16	94.49 15	93.30 21	97.03 14	97.31 10	95.96 13	91.30 19	93.41 19	88.55 25	93.00 20	90.33 29	91.43 27	95.53 7	94.41 14	95.53 58	97.47 15

GBi-Net		84.51 95	84.80 91	84.17 101	84.20 164	89.95 127	89.70 77	80.37 117	81.17 92	75.50 92	69.63 118	79.69 84	79.75 122	90.73 89	90.72 70	95.52 59	91.71 129

test1		84.51 95	84.80 91	84.17 101	84.20 164	89.95 127	89.70 77	80.37 117	81.17 92	75.50 92	69.63 118	79.69 84	79.75 122	90.73 89	90.72 70	95.52 59	91.71 129

FMVSNet2		83.87 100	83.73 102	84.05 105	84.20 164	89.95 127	89.70 77	80.21 122	79.17 113	74.89 97	65.91 139	77.49 96	79.75 122	90.87 86	91.00 66	95.52 59	91.71 129

DPM-MVS		91.72 36	91.48 36	92.00 32	95.53 38	95.75 46	95.94 15	91.07 22	91.20 35	85.58 42	81.63 58	90.74 27	88.40 48	93.40 44	93.75 25	95.45 62	93.85 85

OPM-MVS		87.56 70	85.80 86	89.62 51	93.90 54	94.09 74	94.12 37	88.18 40	75.40 133	77.30 90	76.41 82	77.93 94	88.79 43	92.20 62	90.82 69	95.40 63	93.72 89
Ray L. Khuboni, Hongjun Xu: Octagram Propagation Matching for Multi-Scale View Stereopsis (OPM-MVS).
abl_6				90.66 41	94.65 48	96.27 38	92.21 51	86.94 49	90.23 42	86.38 38	85.50 41	92.96 10	88.37 49			95.40 63	95.46 57

CSCG		92.76 27	93.16 28	92.29 30	96.30 29	97.74 7	94.67 34	88.98 37	92.46 24	89.73 21	86.67 38	92.15 18	88.69 45	92.26 60	92.92 46	95.40 63	97.89 10

canonicalmvs		89.36 52	89.92 47	88.70 60	91.38 79	95.92 44	91.81 58	82.61 98	90.37 41	82.73 59	82.09 52	79.28 87	88.30 50	91.17 76	93.59 28	95.36 66	97.04 26

IB-MVS	79.09 12	82.60 112	82.19 110	83.07 115	91.08 83	93.55 81	80.90 180	81.35 108	76.56 125	80.87 67	64.81 152	69.97 133	68.87 181	85.64 155	90.06 86	95.36 66	94.74 69
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
gg-mvs-nofinetune		75.64 186	77.26 165	73.76 191	87.92 120	92.20 103	87.32 116	64.67 209	51.92 215	35.35 219	46.44 207	77.05 99	71.97 171	92.64 56	91.02 65	95.34 68	89.53 152

Vis-MVSNet		84.38 99	86.68 79	81.70 128	87.65 125	94.89 62	88.14 106	80.90 112	74.48 139	68.23 131	77.53 78	80.72 75	69.98 178	92.68 55	91.90 55	95.33 69	94.58 72
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
TAPA-MVS	84.37 7	88.91 56	88.93 56	88.89 57	93.00 64	94.85 63	92.00 54	84.84 62	91.68 33	80.05 72	79.77 66	84.56 58	88.17 51	90.11 97	89.00 115	95.30 70	92.57 113
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MVS_Test		86.93 73	87.24 72	86.56 77	90.10 102	93.47 82	90.31 70	80.12 123	83.55 73	78.12 83	79.58 67	79.80 82	85.45 74	90.17 96	90.59 76	95.29 71	93.53 92

LGP-MVS_train		88.25 64	88.55 57	87.89 69	92.84 67	93.66 79	93.35 43	85.22 60	85.77 60	74.03 102	86.60 39	76.29 101	86.62 68	91.20 74	90.58 77	95.29 71	95.75 50

UniMVSNet_NR-MVSNet		81.87 118	81.33 118	82.50 120	85.31 150	91.30 109	85.70 138	84.25 68	75.89 129	64.21 151	66.95 135	64.65 153	80.22 112	87.07 131	89.18 110	95.27 73	94.29 76

IS_MVSNet		86.18 77	88.18 62	83.85 107	91.02 84	94.72 66	87.48 113	82.46 99	81.05 96	70.28 119	76.98 80	82.20 70	76.65 146	93.97 34	93.38 37	95.18 74	94.97 63

train_agg		92.87 26	93.53 25	92.09 31	96.88 19	95.38 50	95.94 15	90.59 29	90.65 39	83.65 52	94.31 14	91.87 20	90.30 35	93.38 45	92.42 52	95.17 75	96.73 32

NR-MVSNet		80.25 133	79.98 137	80.56 143	85.20 152	90.94 113	85.65 140	83.58 80	75.74 130	61.36 175	65.30 147	56.75 195	72.38 170	88.46 118	88.80 117	95.16 76	93.87 84

PVSNet_Blended_VisFu		87.40 72	87.80 67	86.92 76	92.86 65	95.40 49	88.56 103	83.45 86	79.55 109	82.26 60	74.49 95	84.03 61	79.24 130	92.97 52	91.53 60	95.15 77	96.65 35

EPP-MVSNet		86.55 74	87.76 69	85.15 89	90.52 92	94.41 69	87.24 119	82.32 100	81.79 88	73.60 104	78.57 73	82.41 68	82.07 91	91.23 72	90.39 79	95.14 78	95.48 56

thisisatest0530		85.15 88	85.86 84	84.33 97	89.19 110	92.57 101	87.22 120	80.11 124	82.15 84	74.41 99	78.15 75	73.80 115	79.90 118	90.99 83	89.58 99	95.13 79	93.75 88

TranMVSNet+NR-MVSNet		80.52 130	79.84 139	81.33 134	84.92 159	90.39 119	85.53 143	84.22 70	74.27 142	60.68 180	64.93 151	59.96 176	77.48 141	86.75 139	89.28 105	95.12 80	93.29 93

tttt0517		85.11 89	85.81 85	84.30 98	89.24 108	92.68 97	87.12 124	80.11 124	81.98 85	74.31 101	78.08 76	73.57 117	79.90 118	91.01 82	89.58 99	95.11 81	93.77 87

UniMVSNet (Re)		81.22 126	81.08 121	81.39 132	85.35 149	91.76 107	84.93 147	82.88 89	76.13 128	65.02 148	64.94 150	63.09 159	75.17 154	87.71 126	89.04 113	94.97 82	94.88 65

FC-MVSNet-train		85.18 87	85.31 89	85.03 90	90.67 88	91.62 108	87.66 111	83.61 77	79.75 107	74.37 100	78.69 72	71.21 127	78.91 131	91.23 72	89.96 89	94.96 83	94.69 71

GeoE		84.62 93	83.98 100	85.35 88	89.34 107	92.83 94	88.34 104	78.95 138	79.29 111	77.16 91	68.10 130	74.56 108	83.40 81	89.31 109	89.23 108	94.92 84	94.57 73

tfpn200view9		82.86 108	81.46 115	84.48 94	90.30 99	93.09 88	89.05 92	82.71 92	75.14 134	69.56 122	65.72 141	63.13 157	80.38 111	91.15 78	89.51 101	94.91 85	92.50 117

thres600view7		82.53 114	81.02 122	84.28 99	90.61 90	93.05 89	88.57 102	82.67 94	74.12 144	68.56 130	65.09 149	62.13 168	80.40 110	91.15 78	89.02 114	94.88 86	92.59 111

thres200		82.77 110	81.25 119	84.54 93	90.38 96	93.05 89	89.13 89	82.67 94	74.40 140	69.53 124	65.69 143	63.03 160	80.63 106	91.15 78	89.42 103	94.88 86	92.04 123

Effi-MVS+		85.33 84	85.08 90	85.63 85	89.69 105	93.42 83	89.90 75	80.31 121	79.32 110	72.48 113	73.52 102	74.03 112	86.55 69	90.99 83	89.98 88	94.83 88	94.27 80

thres400		82.68 111	81.15 120	84.47 95	90.52 92	92.89 93	88.95 96	82.71 92	74.33 141	69.22 127	65.31 146	62.61 163	80.63 106	90.96 85	89.50 102	94.79 89	92.45 119

ACMP	83.90 8	88.32 63	88.06 63	88.62 61	92.18 71	93.98 76	91.28 65	85.24 59	86.69 56	81.23 65	85.62 40	75.13 105	87.01 66	89.83 100	89.77 95	94.79 89	95.43 58
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
DU-MVS		81.20 127	80.30 131	82.25 123	84.98 157	90.94 113	85.70 138	83.58 80	75.74 130	64.21 151	65.30 147	59.60 181	80.22 112	86.89 134	89.31 104	94.77 91	94.29 76

UA-Net		86.07 78	87.78 68	84.06 104	92.85 66	95.11 58	87.73 110	84.38 67	73.22 153	73.18 107	79.99 65	89.22 38	71.47 174	93.22 47	93.03 42	94.76 92	90.69 143

MVS_111021_LR		90.14 46	90.89 42	89.26 54	93.23 60	94.05 75	90.43 69	84.65 63	90.16 43	84.52 48	90.14 29	83.80 62	87.99 52	92.50 58	90.92 67	94.74 93	94.70 70

thres100view900		82.55 113	81.01 124	84.34 96	90.30 99	92.27 102	89.04 93	82.77 91	75.14 134	69.56 122	65.72 141	63.13 157	79.62 125	89.97 99	89.26 106	94.73 94	91.61 135

AdaColmap		90.29 44	88.38 60	92.53 27	96.10 32	95.19 55	92.98 47	91.40 18	89.08 48	88.65 24	78.35 74	81.44 72	91.30 29	90.81 88	90.21 82	94.72 95	93.59 91

PCF-MVS	84.60 6	88.66 57	87.75 70	89.73 49	93.06 63	96.02 41	93.22 45	90.00 32	82.44 81	80.02 74	77.96 77	85.16 56	87.36 61	88.54 116	88.54 120	94.72 95	95.61 54
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
DI_MVS_plusplus_trai		86.41 76	85.54 88	87.42 74	89.24 108	93.13 87	92.16 53	82.65 96	82.30 82	80.75 70	68.30 129	80.41 76	85.01 75	90.56 94	90.07 85	94.70 97	94.01 82

Fast-Effi-MVS+		83.77 102	82.98 105	84.69 91	87.98 119	91.87 106	88.10 107	77.70 152	78.10 119	73.04 109	69.13 124	68.51 141	86.66 67	90.49 95	89.85 93	94.67 98	92.88 100

WR-MVS_H		75.84 184	76.93 170	74.57 190	82.86 183	89.50 143	78.34 193	79.36 135	66.90 182	52.51 198	60.20 174	59.71 178	59.73 200	83.61 178	85.77 158	94.65 99	92.84 101

OMC-MVS		90.23 45	90.40 46	90.03 46	93.45 58	95.29 51	91.89 57	86.34 53	93.25 21	84.94 46	81.72 56	86.65 51	88.90 41	91.69 68	90.27 81	94.65 99	93.95 83

HQP-MVS		89.13 55	89.58 53	88.60 62	93.53 57	93.67 78	93.29 44	87.58 46	88.53 50	75.50 92	87.60 35	80.32 77	87.07 64	90.66 93	89.95 90	94.62 101	96.35 42

PEN-MVS		76.02 181	76.07 177	75.95 180	83.17 177	87.97 161	79.65 184	80.07 127	66.57 184	51.45 201	60.94 168	55.47 200	66.81 190	82.72 183	86.80 139	94.59 102	92.03 124

ACMM	83.27 10	87.68 69	86.09 82	89.54 52	93.26 59	92.19 104	91.43 61	86.74 50	86.02 59	82.85 57	75.63 87	75.14 104	88.41 47	90.68 92	89.99 87	94.59 102	92.97 98
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
CANet_DTU		85.43 83	87.72 71	82.76 118	90.95 87	93.01 91	89.99 73	75.46 169	82.67 75	64.91 149	83.14 46	80.09 79	80.68 104	92.03 66	91.03 64	94.57 104	92.08 121

CP-MVSNet		76.36 178	76.41 174	76.32 177	82.73 186	88.64 156	79.39 187	79.62 130	67.21 180	53.70 195	60.72 170	55.22 201	67.91 186	83.52 179	86.34 149	94.55 105	93.19 94

EG-PatchMatch MVS		76.40 177	75.47 186	77.48 167	85.86 142	90.22 123	82.45 166	73.96 175	59.64 207	59.60 184	52.75 199	62.20 167	68.44 183	88.23 120	87.50 128	94.55 105	87.78 167

tfpnnormal		77.46 165	74.86 190	80.49 144	86.34 138	88.92 154	84.33 154	81.26 109	61.39 202	61.70 172	51.99 201	53.66 207	74.84 157	88.63 115	87.38 131	94.50 107	92.08 121

baseline2		82.80 109	82.86 107	82.73 119	87.68 124	90.50 118	84.92 148	78.93 139	78.07 120	73.06 108	75.08 91	69.77 134	77.31 142	88.90 113	86.94 137	94.50 107	90.74 142

Vis-MVSNet (Re-imp)		83.65 103	86.81 77	79.96 148	90.46 95	92.71 95	84.84 149	82.00 102	80.93 98	62.44 164	76.29 83	82.32 69	65.54 194	92.29 59	91.66 57	94.49 109	91.47 138

PS-CasMVS		75.90 183	75.86 182	75.96 179	82.59 187	88.46 159	79.23 190	79.56 132	66.00 187	52.77 197	59.48 178	54.35 205	67.14 189	83.37 180	86.23 150	94.47 110	93.10 96

Baseline_NR-MVSNet		79.84 137	78.37 154	81.55 131	84.98 157	86.66 171	85.06 145	83.49 82	75.57 132	63.31 158	58.22 186	60.97 171	78.00 137	86.89 134	87.13 133	94.47 110	93.15 95

FMVSNet1		81.64 123	80.61 127	82.84 117	82.36 189	89.20 148	88.67 98	79.58 131	70.79 165	72.63 112	58.95 182	72.26 124	79.34 128	90.73 89	90.72 70	94.47 110	91.62 134

DTE-MVSNet		75.14 188	75.44 187	74.80 187	83.18 176	87.19 168	78.25 195	80.11 124	66.05 186	48.31 206	60.88 169	54.67 202	64.54 195	82.57 185	86.17 151	94.43 113	90.53 147

PLC	83.76 9	88.61 59	86.83 76	90.70 40	94.22 50	92.63 98	91.50 60	87.19 48	89.16 47	86.87 34	75.51 88	80.87 74	89.98 38	90.01 98	89.20 109	94.41 114	90.45 148
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
UGNet		85.90 81	88.23 61	83.18 114	88.96 112	94.10 73	87.52 112	83.60 78	81.66 89	77.90 86	80.76 62	83.19 64	66.70 191	91.13 81	90.71 73	94.39 115	96.06 46
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
TransMVSNet (Re)		76.57 172	75.16 189	78.22 164	85.60 146	87.24 167	82.46 165	81.23 110	59.80 206	59.05 188	57.07 188	59.14 185	66.60 192	88.09 121	86.82 138	94.37 116	87.95 166

gm-plane-assit		70.29 199	70.65 201	69.88 200	85.03 155	78.50 210	58.41 217	65.47 205	50.39 217	40.88 215	49.60 203	50.11 211	75.14 155	91.43 71	89.78 94	94.32 117	84.73 186

CNLPA		88.40 60	87.00 74	90.03 46	93.73 56	94.28 70	89.56 81	85.81 55	91.87 30	87.55 30	69.53 122	81.49 71	89.23 39	89.45 107	88.59 119	94.31 118	93.82 86

MAR-MVS		88.39 62	88.44 59	88.33 66	94.90 43	95.06 59	90.51 68	83.59 79	85.27 63	79.07 79	77.13 79	82.89 66	87.70 54	92.19 63	92.32 53	94.23 119	94.20 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
MSDG		83.87 100	81.02 122	87.19 75	92.17 72	89.80 134	89.15 88	85.72 56	80.61 101	79.24 78	66.66 136	68.75 140	82.69 84	87.95 123	87.44 129	94.19 120	85.92 180

pm-mvs1		78.51 157	77.75 162	79.40 151	84.83 160	89.30 145	83.55 160	79.38 134	62.64 198	63.68 156	58.73 184	64.68 152	70.78 177	89.79 101	87.84 125	94.17 121	91.28 140

CPTT-MVS		91.39 38	90.95 41	91.91 33	95.06 40	95.24 54	95.02 31	88.98 37	91.02 36	86.71 35	84.89 44	88.58 44	91.60 23	90.82 87	89.67 98	94.08 122	96.45 37

v144192		78.81 151	77.22 166	80.67 141	82.95 180	89.79 135	86.40 131	77.42 153	68.26 179	63.13 159	59.50 177	58.13 187	80.08 117	85.93 151	86.08 153	94.06 123	92.83 102

diffmvs		86.52 75	86.76 78	86.23 80	88.31 118	92.63 98	89.58 80	81.61 105	86.14 58	80.26 71	79.00 70	77.27 97	83.58 79	88.94 112	89.06 112	94.05 124	94.29 76

v1921920		78.57 156	76.99 169	80.41 146	82.93 181	89.63 141	86.38 132	77.14 156	68.31 178	61.80 171	58.89 183	56.79 194	80.19 115	86.50 146	86.05 155	94.02 125	92.76 105

ACMH	78.52 14	81.86 119	80.45 130	83.51 113	90.51 94	91.22 110	85.62 141	84.23 69	70.29 170	62.21 165	69.04 126	64.05 155	84.48 77	87.57 127	88.45 122	94.01 126	92.54 115
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
WR-MVS		76.63 171	78.02 159	75.02 185	84.14 167	89.76 136	78.34 193	80.64 115	69.56 171	52.32 199	61.26 164	61.24 170	60.66 199	84.45 173	87.07 134	93.99 127	92.77 104

v10		79.62 140	78.19 155	81.28 135	83.73 170	89.69 138	87.27 118	76.86 159	70.50 168	65.46 143	60.58 172	60.47 173	80.44 109	86.91 133	86.63 143	93.93 128	92.55 114

anonymousdsp		77.94 160	79.00 146	76.71 173	79.03 201	87.83 162	79.58 185	72.87 177	65.80 189	58.86 189	65.82 140	62.48 165	75.99 149	86.77 138	88.66 118	93.92 129	95.68 53

v1240		78.15 158	76.53 172	80.04 147	82.85 184	89.48 144	85.61 142	76.77 160	67.05 181	61.18 178	58.37 185	56.16 198	79.89 120	86.11 150	86.08 153	93.92 129	92.47 118

v8		79.90 136	78.39 153	81.66 129	83.97 168	89.81 133	87.16 122	77.40 154	71.49 159	67.71 132	61.24 165	62.49 164	79.83 121	85.48 159	86.17 151	93.89 131	92.02 125

v2v482		79.84 137	78.07 157	81.90 126	83.75 169	90.21 124	87.17 121	79.85 129	70.65 166	65.93 141	61.93 162	60.07 175	80.82 101	85.25 161	86.71 140	93.88 132	91.70 133

thisisatest0515		79.76 139	80.59 128	78.80 156	84.40 162	88.91 155	79.48 186	76.94 158	72.29 157	67.33 134	67.82 132	65.99 148	70.80 176	88.50 117	87.84 125	93.86 133	92.75 106

IterMVS-LS		83.28 106	82.95 106	83.65 108	88.39 117	88.63 157	86.80 128	78.64 143	76.56 125	73.43 106	72.52 107	75.35 103	80.81 102	86.43 147	88.51 121	93.84 134	92.66 108
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
TSAR-MVS + ACMM		92.97 24	94.51 14	91.16 38	95.88 35	96.59 31	95.09 30	90.45 31	93.42 18	83.01 55	94.68 11	90.74 27	88.74 44	94.75 19	93.78 24	93.82 135	97.63 12

v1144		79.38 146	77.83 160	81.18 136	83.62 171	90.23 122	87.15 123	78.35 145	69.13 173	64.02 154	60.20 174	59.41 182	80.14 116	86.78 137	86.57 144	93.81 136	92.53 116

v1192		78.94 150	77.33 164	80.82 139	83.25 175	89.90 131	86.91 126	77.72 151	68.63 177	62.61 163	59.17 179	57.53 191	80.62 108	86.89 134	86.47 146	93.79 137	92.75 106

TSAR-MVS + COLMAP		88.40 60	89.09 55	87.60 73	92.72 68	93.92 77	92.21 51	85.57 57	91.73 31	73.72 103	91.75 24	73.22 121	87.64 57	91.49 70	89.71 97	93.73 138	91.82 127

UniMVSNet_ETH3D		79.24 147	76.47 173	82.48 121	85.66 145	90.97 112	86.08 135	81.63 104	64.48 194	68.94 129	54.47 194	57.65 190	78.83 132	85.20 165	88.91 116	93.72 139	93.60 90

EPNet		89.60 49	89.91 48	89.24 55	96.45 28	93.61 80	92.95 48	88.03 41	85.74 61	83.36 53	87.29 37	83.05 65	80.98 100	92.22 61	91.85 56	93.69 140	95.58 55
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
HyFIR lowres test		81.62 125	79.45 145	84.14 103	91.00 85	93.38 84	88.27 105	78.19 146	76.28 127	70.18 120	48.78 204	73.69 116	83.52 80	87.05 132	87.83 127	93.68 141	89.15 154

V42		79.59 141	78.43 152	80.94 138	82.79 185	89.71 137	86.66 129	76.73 161	71.38 160	67.42 133	61.01 167	62.30 166	78.39 134	85.56 157	86.48 145	93.65 142	92.60 110

DCV-MVSNet		85.88 82	86.17 80	85.54 87	89.10 111	89.85 132	89.34 84	80.70 114	83.04 74	78.08 85	76.19 84	79.00 88	82.42 88	89.67 103	90.30 80	93.63 143	95.12 60

v7n		77.22 166	76.23 176	78.38 163	81.89 192	89.10 152	82.24 171	76.36 162	65.96 188	61.21 177	56.56 189	55.79 199	75.07 156	86.55 143	86.68 141	93.52 144	92.95 99

Anonymous20231211		84.42 98	83.02 104	86.05 82	88.85 113	92.70 96	88.92 97	83.40 87	79.99 104	78.31 82	55.83 191	78.92 89	83.33 82	89.06 111	89.76 96	93.50 145	94.90 64

ACMH+	79.08 13	81.84 120	80.06 135	83.91 106	89.92 104	90.62 115	86.21 133	83.48 84	73.88 146	65.75 142	66.38 137	65.30 151	84.63 76	85.90 152	87.25 132	93.45 146	91.13 141

pmmvs5		76.93 168	76.33 175	77.62 166	81.97 191	88.40 160	81.32 176	74.35 173	65.42 192	61.42 174	63.07 158	57.95 189	73.23 168	85.60 156	85.35 163	93.41 147	88.55 158

Anonymous202405211			82.75 108		89.58 106	92.97 92	89.04 93	84.13 71	78.72 115		57.18 187	76.64 100	83.13 83	89.55 105	89.92 91	93.38 148	94.28 79

USDC		80.69 129	79.89 138	81.62 130	86.48 136	89.11 151	86.53 130	78.86 140	81.15 95	63.48 157	72.98 104	59.12 186	81.16 98	87.10 130	85.01 165	93.23 149	84.77 185

LS3D		85.96 80	84.37 96	87.81 70	94.13 51	93.27 85	90.26 72	89.00 35	84.91 68	72.84 111	71.74 108	72.47 123	87.45 60	89.53 106	89.09 111	93.20 150	89.60 151

MS-PatchMatch		81.79 121	81.44 116	82.19 125	90.35 97	89.29 146	88.08 108	75.36 170	77.60 121	69.00 128	64.37 155	78.87 90	77.14 145	88.03 122	85.70 159	93.19 151	86.24 177

Fast-Effi-MVS+-dtu		79.95 135	80.69 126	79.08 153	86.36 137	89.14 150	85.85 136	72.28 179	72.85 156	59.32 185	70.43 116	68.42 143	77.57 140	86.14 149	86.44 147	93.11 152	91.39 139

GA-MVS		79.52 142	79.71 142	79.30 152	85.68 144	90.36 120	84.55 151	78.44 144	70.47 169	57.87 190	68.52 128	61.38 169	76.21 148	89.40 108	87.89 124	93.04 153	89.96 150

DeepPCF-MVS	88.51 2	92.64 30	94.42 17	90.56 42	94.84 45	96.92 19	91.31 64	89.61 33	95.16 5	84.55 47	89.91 30	91.45 22	90.15 37	95.12 11	94.81 7	92.90 154	97.58 13

CDS-MVSNet		81.63 124	82.09 111	81.09 137	87.21 130	90.28 121	87.46 115	80.33 120	69.06 174	70.66 116	71.30 109	73.87 113	67.99 184	89.58 104	89.87 92	92.87 155	90.69 143
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
Effi-MVS+-dtu		82.05 116	81.76 112	82.38 122	87.72 122	90.56 117	86.90 127	78.05 148	73.85 147	66.85 136	71.29 110	71.90 125	82.00 92	86.64 142	85.48 161	92.76 156	92.58 112

v148		78.59 155	76.84 171	80.62 142	83.61 172	89.16 149	83.65 159	79.24 136	69.38 172	69.34 126	59.88 176	60.41 174	75.19 153	83.81 177	84.63 170	92.70 157	90.63 145

pmmvs4		79.99 134	78.08 156	82.22 124	83.04 179	87.16 169	84.95 146	78.80 142	78.64 116	74.53 98	64.61 153	59.41 182	79.45 127	84.13 175	84.54 172	92.53 158	88.08 163

test_part1		83.23 107	80.55 129	86.35 79	88.60 115	90.61 116	90.78 67	81.13 111	70.89 164	83.01 55	55.72 192	74.60 107	82.19 89	87.79 124	89.26 106	92.39 159	95.01 61

RPMNet		77.07 167	77.63 163	76.42 175	85.56 147	85.15 184	81.37 174	65.27 206	74.71 137	60.29 181	63.71 157	66.59 147	73.64 164	82.71 184	82.12 186	92.38 160	88.39 159

CR-MVSNet		78.71 153	78.86 147	78.55 160	85.85 143	85.15 184	82.30 169	68.23 195	74.71 137	65.37 145	64.39 154	69.59 136	77.18 143	85.10 167	84.87 166	92.34 161	88.21 161

COLMAP_ROB	76.78 15	80.50 131	78.49 150	82.85 116	90.96 86	89.65 140	86.20 134	83.40 87	77.15 123	66.54 137	62.27 160	65.62 150	77.89 138	85.23 162	84.70 169	92.11 162	84.83 184
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
PatchMatch-RL		83.34 105	81.36 117	85.65 84	90.33 98	89.52 142	84.36 153	81.82 103	80.87 100	79.29 77	74.04 97	62.85 162	86.05 71	88.40 119	87.04 136	92.04 163	86.77 173

PMMVS		81.65 122	84.05 99	78.86 155	78.56 203	82.63 197	83.10 161	67.22 199	81.39 90	70.11 121	84.91 43	79.74 83	82.12 90	87.31 128	85.70 159	92.03 164	86.67 176

IterMVS-SCA-FT		79.41 145	80.20 133	78.49 161	85.88 140	86.26 173	83.95 156	71.94 180	73.55 151	61.94 168	70.48 115	70.50 130	75.23 152	85.81 154	84.61 171	91.99 165	90.18 149

baseline		84.89 90	86.06 83	83.52 112	87.25 129	89.67 139	87.76 109	75.68 168	84.92 67	78.40 81	80.10 63	80.98 73	80.20 114	86.69 141	87.05 135	91.86 166	92.99 97

MIMVSNet		74.69 190	75.60 185	73.62 192	76.02 210	85.31 183	81.21 179	67.43 198	71.02 162	59.07 187	54.48 193	64.07 154	66.14 193	86.52 145	86.64 142	91.83 167	81.17 198

FC-MVSNet-test		76.53 174	81.62 114	70.58 199	84.99 156	85.73 178	74.81 201	78.85 141	77.00 124	39.13 217	75.90 85	73.50 118	54.08 206	86.54 144	85.99 156	91.65 168	86.68 174

pmmvs-eth3d		74.32 192	71.96 198	77.08 170	77.33 206	82.71 196	78.41 192	76.02 166	66.65 183	65.98 140	54.23 196	49.02 214	73.14 169	82.37 187	82.69 183	91.61 169	86.05 179

SixPastTwentyTwo		76.02 181	75.72 183	76.36 176	83.38 173	87.54 164	75.50 200	76.22 163	65.50 191	57.05 191	70.64 112	53.97 206	74.54 159	80.96 191	82.12 186	91.44 170	89.35 153

pmmvs6		74.83 189	72.89 196	77.09 169	82.11 190	87.50 165	80.88 181	76.97 157	52.79 214	61.91 170	46.66 206	60.49 172	69.28 180	86.74 140	85.46 162	91.39 171	90.56 146

test-mter		77.79 161	80.02 136	75.18 184	81.18 197	82.85 195	80.52 183	62.03 213	73.62 150	62.16 166	73.55 101	73.83 114	73.81 163	84.67 170	83.34 178	91.37 172	88.31 160

TDRefinement		79.05 149	77.05 168	81.39 132	88.45 116	89.00 153	86.92 125	82.65 96	74.21 143	64.41 150	59.17 179	59.16 184	74.52 160	85.23 162	85.09 164	91.37 172	87.51 169

test-LLR		79.47 144	79.84 139	79.03 154	87.47 126	82.40 200	81.24 177	78.05 148	73.72 148	62.69 161	73.76 99	74.42 109	73.49 165	84.61 171	82.99 181	91.25 174	87.01 171

TESTMET0.1,1		77.78 162	79.84 139	75.38 183	80.86 198	82.40 200	81.24 177	62.72 212	73.72 148	62.69 161	73.76 99	74.42 109	73.49 165	84.61 171	82.99 181	91.25 174	87.01 171

TinyColmap		76.73 169	73.95 193	79.96 148	85.16 154	85.64 180	82.34 168	78.19 146	70.63 167	62.06 167	60.69 171	49.61 212	80.81 102	85.12 166	83.69 177	91.22 176	82.27 193

FMVSNet5		75.50 187	76.07 177	74.83 186	76.16 208	81.19 203	81.34 175	70.21 188	73.20 154	61.59 173	58.97 181	68.33 144	68.50 182	85.87 153	85.85 157	91.18 177	79.11 204

IterMVS		78.79 152	79.71 142	77.71 165	85.26 151	85.91 176	84.54 152	69.84 191	73.38 152	61.25 176	70.53 114	70.35 131	74.43 161	85.21 164	83.80 176	90.95 178	88.77 156
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
PatchT		76.42 175	77.81 161	74.80 187	78.46 204	84.30 189	71.82 207	65.03 208	73.89 145	65.37 145	61.58 163	66.70 146	77.18 143	85.10 167	84.87 166	90.94 179	88.21 161

EPNet_dtu		81.98 117	83.82 101	79.83 150	94.10 52	85.97 175	87.29 117	84.08 72	80.61 101	59.96 182	81.62 59	77.19 98	62.91 198	87.21 129	86.38 148	90.66 180	87.77 168
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
PM-MVS		74.17 193	73.10 194	75.41 182	76.07 209	82.53 198	77.56 196	71.69 181	71.04 161	61.92 169	61.23 166	47.30 215	74.82 158	81.78 189	79.80 190	90.42 181	88.05 164

test0.0.03 1		76.03 180	78.51 149	73.12 195	87.47 126	85.13 186	76.32 198	78.05 148	73.19 155	50.98 204	70.64 112	69.28 137	55.53 202	85.33 160	84.38 173	90.39 182	81.63 196

Anonymous20231206		70.80 198	70.59 202	71.04 198	81.60 194	82.49 199	74.64 202	75.87 167	64.17 195	49.27 205	44.85 210	53.59 208	54.68 205	83.07 181	82.34 185	90.17 183	83.65 188

CHOSEN 1792x2688		82.16 115	80.91 125	83.61 109	91.14 82	92.01 105	89.55 82	79.15 137	79.87 105	70.29 118	52.51 200	72.56 122	81.39 95	88.87 114	88.17 123	90.15 184	92.37 120

MIMVSNet1		65.00 205	66.24 206	63.55 208	58.41 219	80.01 207	69.00 210	74.03 174	55.81 212	41.88 214	36.81 215	49.48 213	47.89 211	81.32 190	82.40 184	90.08 185	77.88 206

LTVRE_ROB	74.41 16	75.78 185	74.72 191	77.02 171	85.88 140	89.22 147	82.44 167	77.17 155	50.57 216	45.45 210	65.44 145	52.29 209	81.25 96	85.50 158	87.42 130	89.94 186	92.62 109
Andreas Kuhn, Heiko HirschmÃ¼ller, Daniel Scharstein, Helmut Mayer: A TV Prior for High-Quality Scalable Multi-View Stereo Reconstruction. International Journal of Computer Vision 2016
GG-mvs-BLEND		57.56 210	82.61 109	28.34 218	0.22 226	90.10 126	79.37 188	0.14 224	79.56 108	0.40 227	71.25 111	83.40 63	0.30 224	86.27 148	83.87 174	89.59 187	83.83 187

CVMVSNet		76.70 170	78.46 151	74.64 189	83.34 174	84.48 188	81.83 173	74.58 171	68.88 175	51.23 203	69.77 117	70.05 132	67.49 187	84.27 174	83.81 175	89.38 188	87.96 165

TAMVS		76.42 175	77.16 167	75.56 181	83.05 178	85.55 181	80.58 182	71.43 182	65.40 193	61.04 179	67.27 134	69.22 139	67.99 184	84.88 169	84.78 168	89.28 189	83.01 191

CostFormer		80.94 128	80.21 132	81.79 127	87.69 123	88.58 158	87.47 114	70.66 185	80.02 103	77.88 87	73.03 103	71.40 126	78.24 135	79.96 195	79.63 191	88.82 190	88.84 155

RPSCF		83.46 104	83.36 103	83.59 110	87.75 121	87.35 166	84.82 150	79.46 133	83.84 72	78.12 83	82.69 50	79.87 80	82.60 87	82.47 186	81.13 189	88.78 191	86.13 178

test20.03		68.31 202	70.05 203	66.28 206	82.41 188	80.84 204	67.35 211	76.11 165	58.44 209	40.80 216	53.77 197	54.54 203	42.28 213	83.07 181	81.96 188	88.73 192	77.76 207

SCA		79.51 143	80.15 134	78.75 157	86.58 135	87.70 163	83.07 162	68.53 194	81.31 91	66.40 138	73.83 98	75.38 102	79.30 129	80.49 193	79.39 194	88.63 193	82.96 192

testgi		71.92 197	74.20 192	69.27 201	84.58 161	83.06 192	73.40 204	74.39 172	64.04 196	46.17 209	68.90 127	57.15 193	48.89 210	84.07 176	83.08 180	88.18 194	79.09 205

dps		78.02 159	75.94 181	80.44 145	86.06 139	86.62 172	82.58 164	69.98 189	75.14 134	77.76 89	69.08 125	59.93 177	78.47 133	79.47 197	77.96 198	87.78 195	83.40 189

PatchmatchNet		78.67 154	78.85 148	78.46 162	86.85 134	86.03 174	83.77 158	68.11 197	80.88 99	66.19 139	72.90 105	73.40 119	78.06 136	79.25 199	77.71 199	87.75 196	81.75 195
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
MDTV_nov1_ep13_2view		73.21 195	72.91 195	73.56 193	80.01 199	84.28 190	78.62 191	66.43 203	68.64 176	59.12 186	60.39 173	59.69 180	69.81 179	78.82 201	77.43 200	87.36 197	81.11 199

MDTV_nov1_ep13		79.14 148	79.49 144	78.74 158	85.40 148	86.89 170	84.32 155	70.29 187	78.85 114	69.42 125	75.37 89	73.29 120	75.64 151	80.61 192	79.48 193	87.36 197	81.91 194

EPMVS		77.53 164	78.07 157	76.90 172	86.89 133	84.91 187	82.18 172	66.64 202	81.00 97	64.11 153	72.75 106	69.68 135	74.42 162	79.36 198	78.13 197	87.14 199	80.68 201

EU-MVSNet		69.98 200	72.30 197	67.28 204	75.67 211	79.39 208	73.12 205	69.94 190	63.59 197	42.80 213	62.93 159	56.71 196	55.07 204	79.13 200	78.55 196	87.06 200	85.82 181

new-patchmatchnet		63.80 206	63.31 208	64.37 207	76.49 207	75.99 211	63.73 214	70.99 184	57.27 210	43.08 212	45.86 208	43.80 216	45.13 212	73.20 209	70.68 212	86.80 201	76.34 209

pmnet_mix02		71.95 196	71.83 199	72.10 196	81.40 196	80.63 206	73.78 203	72.85 178	70.90 163	54.89 193	62.17 161	57.42 192	62.92 197	76.80 204	73.98 208	86.74 202	80.87 200

CMPMVS	56.49 17	73.84 194	71.73 200	76.31 178	85.20 152	85.67 179	75.80 199	73.23 176	62.26 199	65.40 144	53.40 198	59.70 179	71.77 173	80.25 194	79.56 192	86.45 203	81.28 197
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
MDA-MVSNet-bldmvs		66.22 204	64.49 207	68.24 202	61.67 216	82.11 202	70.07 209	76.16 164	59.14 208	47.94 207	54.35 195	35.82 222	67.33 188	64.94 214	75.68 203	86.30 204	79.36 203

MVS-HIRNet		68.83 201	66.39 205	71.68 197	77.58 205	75.52 212	66.45 212	65.05 207	62.16 200	62.84 160	44.76 211	56.60 197	71.96 172	78.04 202	75.06 206	86.18 205	72.56 211

CHOSEN 280x420		80.28 132	81.66 113	78.67 159	82.92 182	79.24 209	85.36 144	66.79 201	78.11 118	70.32 117	75.03 92	79.87 80	81.09 99	89.07 110	83.16 179	85.54 206	87.17 170

tpm cat1		77.78 162	75.28 188	80.70 140	87.14 131	85.84 177	85.81 137	70.40 186	77.44 122	78.80 80	63.72 156	64.01 156	76.55 147	75.60 207	75.21 205	85.51 207	85.12 182

tpm		76.30 179	76.05 179	76.59 174	86.97 132	83.01 194	83.83 157	67.06 200	71.83 158	63.87 155	69.56 121	62.88 161	73.41 167	79.79 196	78.59 195	84.41 208	86.68 174

tpmrst		76.55 173	75.99 180	77.20 168	87.32 128	83.05 193	82.86 163	65.62 204	78.61 117	67.22 135	69.19 123	65.71 149	75.87 150	76.75 205	75.33 204	84.31 209	83.28 190

pmmvs3		61.89 208	61.74 210	62.06 209	64.30 215	70.83 216	64.22 213	52.14 217	48.78 218	44.47 211	41.67 213	41.70 220	63.03 196	76.06 206	76.02 202	84.18 210	77.14 208

ADS-MVSNet		74.53 191	75.69 184	73.17 194	81.57 195	80.71 205	79.27 189	63.03 211	79.27 112	59.94 183	67.86 131	68.32 145	71.08 175	77.33 203	76.83 201	84.12 211	79.53 202

ambc			61.92 209		70.98 214	73.54 214	63.64 215		60.06 204	52.23 200	38.44 214	19.17 225	57.12 201	82.33 188	75.03 207	83.21 212	84.89 183

FPMVS		63.63 207	60.08 212	67.78 203	80.01 199	71.50 215	72.88 206	69.41 193	61.82 201	53.11 196	45.12 209	42.11 219	50.86 208	66.69 212	63.84 213	80.41 213	69.46 213

N_pmnet		66.85 203	66.63 204	67.11 205	78.73 202	74.66 213	70.53 208	71.07 183	66.46 185	46.54 208	51.68 202	51.91 210	55.48 203	74.68 208	72.38 209	80.29 214	74.65 210

new_pmnet		59.28 209	61.47 211	56.73 211	61.66 217	68.29 217	59.57 216	54.91 214	60.83 203	34.38 220	44.66 212	43.65 217	49.90 209	71.66 210	71.56 211	79.94 215	69.67 212

PMVS	50.48 18	55.81 211	51.93 213	60.33 210	72.90 213	49.34 219	48.78 218	69.51 192	43.49 219	54.25 194	36.26 216	41.04 221	39.71 215	65.07 213	60.70 214	76.85 216	67.58 214
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
Gipuma		49.17 212	47.05 215	51.65 212	59.67 218	48.39 220	41.98 221	63.47 210	55.64 213	33.33 221	14.90 219	13.78 226	41.34 214	69.31 211	72.30 210	70.11 217	55.00 218
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PMMVS2		41.68 214	44.74 216	38.10 213	46.97 222	52.32 218	40.63 222	48.08 218	35.51 220	7.36 226	26.86 218	24.64 224	16.72 220	55.24 217	59.03 215	68.85 218	59.59 217

test_method		41.78 213	48.10 214	34.42 216	10.74 225	19.78 226	44.64 220	17.73 221	59.83 205	38.67 218	35.82 217	54.41 204	34.94 216	62.87 215	43.13 218	59.81 219	60.82 216

DeepMVS_CX						48.31 221	48.03 219	26.08 220	56.42 211	25.77 222	47.51 205	31.31 223	51.30 207	48.49 218		53.61 220	61.52 215

tmp_tt				32.73 217	43.96 223	21.15 225	26.71 223	8.99 222	65.67 190	51.39 202	56.01 190	42.64 218	11.76 221	56.60 216	50.81 217	53.55 221

E-PMN		31.40 215	26.80 218	36.78 214	51.39 221	29.96 223	20.20 224	54.17 215	25.93 222	12.75 224	14.73 220	8.58 228	34.10 218	27.36 220	37.83 219	48.07 222	43.18 220

EMVS		30.49 217	25.44 219	36.39 215	51.47 220	29.89 224	20.17 225	54.00 216	26.49 221	12.02 225	13.94 222	8.84 227	34.37 217	25.04 221	34.37 220	46.29 223	39.53 221

MVE	30.17 19	30.88 216	33.52 217	27.80 219	23.78 224	39.16 222	18.69 226	46.90 219	21.88 223	15.39 223	14.37 221	7.31 229	24.41 219	41.63 219	56.22 216	37.64 224	54.07 219
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
testmvs		1.03 218	1.63 220	0.34 220	0.09 227	0.35 227	0.61 228	0.16 223	1.49 224	0.10 228	3.15 223	0.15 230	0.86 223	1.32 222	1.18 221	0.20 225	3.76 223

test123		0.87 219	1.40 221	0.25 221	0.03 228	0.25 228	0.35 229	0.08 225	1.21 225	0.05 229	2.84 224	0.03 231	0.89 222	0.43 223	1.16 222	0.13 226	3.87 222

uanet_test		0.00 220	0.00 222	0.00 222	0.00 229	0.00 229	0.00 230	0.00 226	0.00 226	0.00 230	0.00 225	0.00 232	0.00 225	0.00 224	0.00 223	0.00 227	0.00 224

sosnet-low-res		0.00 220	0.00 222	0.00 222	0.00 229	0.00 229	0.00 230	0.00 226	0.00 226	0.00 230	0.00 225	0.00 232	0.00 225	0.00 224	0.00 223	0.00 227	0.00 224

sosnet		0.00 220	0.00 222	0.00 222	0.00 229	0.00 229	0.00 230	0.00 226	0.00 226	0.00 230	0.00 225	0.00 232	0.00 225	0.00 224	0.00 223	0.00 227	0.00 224

RE-MVS-def										56.08 192

9.14												92.16 17

SR-MVS					96.58 27			90.99 23				92.40 14

our_test_3					81.81 193	83.96 191	76.61 197

MTAPA										92.97 2		91.03 23

MTMP										93.14 1		90.21 31

Patchmatch-RL test							8.55 227

mPP-MVS					97.06 13							88.08 46

NP-MVS									87.47 54

Patchmtry						85.54 182	82.30 169	68.23 195		65.37 145

High-resolution multi-view benchmark