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.

TDRefinement		93.16 1	95.57 1	90.36 1	88.79 52	93.57 1	97.27 1	78.23 21	95.55 1	93.00 1	93.98 17	96.01 38	87.53 1	97.69 1	96.81 1	97.33 1	95.34 4

COLMAP_ROB	85.66 2	91.85 2	95.01 2	88.16 11	88.98 51	92.86 2	95.51 19	72.17 62	94.95 4	91.27 3	94.11 16	97.77 11	84.22 8	96.49 4	95.27 5	96.79 2	93.60 12
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
LTVRE_ROB	86.82 1	91.55 3	94.43 3	88.19 10	83.19 111	86.35 65	93.60 37	78.79 18	95.48 3	91.79 2	93.08 26	97.21 20	86.34 3	97.06 2	96.27 3	95.46 23	95.56 3
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
DVP-MVS++		90.50 10	94.18 4	86.21 27	92.52 7	90.29 28	95.29 22	76.02 41	94.24 5	82.82 55	95.84 5	97.56 15	76.82 55	93.13 38	91.20 44	93.78 45	97.01 1

WR-MVS		89.79 23	93.66 5	85.27 37	91.32 23	88.27 46	93.49 38	79.86 10	92.75 9	75.37 101	96.86 1	98.38 5	75.10 71	95.93 8	94.07 14	96.46 5	89.39 56

WR-MVS_H		88.99 35	93.28 6	83.99 53	91.92 11	89.13 40	91.95 46	83.23 1	90.14 29	71.92 124	95.85 4	98.01 10	71.83 97	95.82 9	93.19 22	93.07 59	90.83 47

PS-CasMVS		89.07 32	93.23 7	84.21 50	92.44 8	88.23 48	90.54 62	82.95 3	90.50 25	75.31 102	95.80 6	98.37 6	71.16 100	96.30 5	93.32 21	92.88 61	90.11 50

ACMH+	79.05 11	89.62 26	93.08 8	85.58 32	88.58 55	89.26 39	92.18 45	74.23 52	93.55 8	82.66 58	92.32 36	98.35 7	80.29 29	95.28 18	92.34 31	95.52 22	90.43 48

PEN-MVS		88.86 39	92.92 9	84.11 52	92.92 4	88.05 51	90.83 55	82.67 5	91.04 18	74.83 104	95.97 3	98.47 3	70.38 107	95.70 13	92.43 30	93.05 60	88.78 62

APDe-MVS		89.85 20	92.91 10	86.29 26	90.47 38	91.34 7	96.04 14	76.41 39	91.11 17	78.50 88	93.44 21	95.82 42	81.55 23	93.16 37	91.90 38	94.77 32	93.58 15

ACMMPR		91.30 4	92.88 11	89.46 4	91.92 11	91.61 5	96.60 5	79.46 14	90.08 30	88.53 13	89.54 65	95.57 47	84.25 7	95.24 20	94.27 12	95.97 11	93.85 8

DTE-MVSNet		88.99 35	92.77 12	84.59 43	93.31 2	88.10 49	90.96 53	83.09 2	91.38 14	76.21 95	96.03 2	98.04 8	70.78 106	95.65 14	92.32 32	93.18 56	87.84 70

DVP-MVS		89.40 27	92.69 13	85.56 34	89.01 50	89.85 33	93.72 35	75.42 45	92.28 11	80.49 72	94.36 13	94.87 65	81.46 24	92.49 49	91.42 41	93.27 53	93.54 17
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
PMVS	79.51 9	90.23 14	92.67 14	87.39 20	90.16 39	88.75 42	93.64 36	75.78 44	90.00 32	83.70 47	92.97 28	92.22 103	86.13 4	97.01 3	96.79 2	94.94 28	90.96 45
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
CP-MVSNet		88.71 41	92.63 15	84.13 51	92.39 9	88.09 50	90.47 66	82.86 4	88.79 41	75.16 103	94.87 9	97.68 13	71.05 102	96.16 6	93.18 23	92.85 62	89.64 54

ACMH	78.40 12	88.94 38	92.62 16	84.65 42	86.45 74	87.16 59	91.47 49	68.79 87	95.49 2	89.74 6	93.55 19	98.50 2	77.96 46	94.14 31	89.57 61	93.49 47	89.94 52
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
RPSCF		88.05 46	92.61 17	82.73 65	84.24 95	88.40 44	90.04 72	66.29 107	91.46 13	82.29 60	88.93 75	96.01 38	79.38 32	95.15 21	94.90 6	94.15 39	93.40 20

DeepC-MVS	83.59 4	90.37 12	92.56 18	87.82 14	91.26 27	92.33 3	94.72 30	80.04 9	90.01 31	84.61 42	93.33 22	94.22 78	80.59 27	92.90 43	92.52 28	95.69 21	92.57 28
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
ACMM	80.67 7	90.67 7	92.46 19	88.57 7	91.35 22	89.93 32	96.34 11	77.36 30	90.17 28	86.88 29	87.32 92	96.63 23	83.32 13	95.79 10	94.49 9	96.19 9	92.91 26
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
ACMMP		90.63 8	92.40 20	88.56 8	91.24 28	91.60 6	96.49 9	77.53 26	87.89 47	86.87 30	87.24 94	96.46 25	82.87 16	95.59 15	94.50 8	96.35 6	93.51 18
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
LGP-MVS_train		90.56 9	92.38 21	88.43 9	90.88 32	91.15 11	95.35 21	77.65 25	86.26 63	87.23 23	90.45 55	97.35 17	83.20 14	95.44 16	93.41 20	96.28 8	92.63 27

SED-MVS		88.96 37	92.37 22	84.99 40	88.64 54	89.65 37	95.11 25	75.98 42	90.73 23	80.15 77	94.21 15	94.51 74	76.59 56	92.94 41	91.17 45	93.46 50	93.37 22

HFP-MVS		90.32 13	92.37 22	87.94 13	91.46 21	90.91 18	95.69 17	79.49 12	89.94 33	83.50 50	89.06 72	94.44 75	81.68 22	94.17 30	94.19 13	95.81 17	93.87 7

DPE-MVS		89.81 22	92.34 24	86.86 23	89.69 44	91.00 16	95.53 18	76.91 33	88.18 45	83.43 53	93.48 20	95.19 57	81.07 26	92.75 45	92.07 36	94.55 36	93.74 11
Kehua Chen, Zhenlong Yuan, Tianlu Mao, Zhaoqi Wang: Dual-Level Precision Edges Guided Multi-View Stereo with Accurate Planarization. AAAI2025
CP-MVS		91.09 5	92.33 25	89.65 2	92.16 10	90.41 27	96.46 10	80.38 8	88.26 44	89.17 10	87.00 97	96.34 30	83.95 10	95.77 11	94.72 7	95.81 17	93.78 10

ACMP	80.00 8	90.12 16	92.30 26	87.58 18	90.83 34	91.10 12	94.96 28	76.06 40	87.47 51	85.33 39	88.91 76	97.65 14	82.13 19	95.31 17	93.44 19	96.14 10	92.22 33
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
SMA-MVS		90.13 15	92.26 27	87.64 17	91.68 16	90.44 26	95.22 24	77.34 32	90.79 22	87.80 16	90.42 56	92.05 108	79.05 35	93.89 32	93.59 18	94.77 32	94.62 5
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
TSAR-MVS + MP.		89.67 24	92.25 28	86.65 25	91.53 18	90.98 17	96.15 13	73.30 56	87.88 48	81.83 66	92.92 29	95.15 60	82.23 18	93.58 34	92.25 33	94.87 29	93.01 25
Zhenlong Yuan, Jiakai Cao, Zhaoqi Wang, Zhaoxin Li: TSAR-MVS: Textureless-aware Segmentation and Correlative Refinement Guided Multi-View Stereo. Pattern Recognition
OPM-MVS		89.82 21	92.24 29	86.99 22	90.86 33	89.35 38	95.07 27	75.91 43	91.16 16	86.87 30	91.07 51	97.29 18	79.13 34	93.32 35	91.99 37	94.12 40	91.49 40
Ray L. Khuboni, Hongjun Xu: Octagram Propagation Matching for Multi-Scale View Stereopsis (OPM-MVS).
SD-MVS		89.91 18	92.23 30	87.19 21	91.31 24	89.79 35	94.31 32	75.34 47	89.26 36	81.79 67	92.68 31	95.08 62	83.88 11	93.10 39	92.69 25	96.54 4	93.02 24
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
SixPastTwentyTwo		89.14 29	92.19 31	85.58 32	84.62 88	82.56 90	90.53 63	71.93 64	91.95 12	85.89 35	94.22 14	97.25 19	85.42 5	95.73 12	91.71 40	95.08 27	91.89 36

TSAR-MVS + ACMM		89.14 29	92.11 32	85.67 31	89.27 47	90.61 24	90.98 52	79.48 13	88.86 39	79.80 79	93.01 27	93.53 87	83.17 15	92.75 45	92.45 29	91.32 82	93.59 13

ACMMP_NAP		89.86 19	91.96 33	87.42 19	91.00 30	90.08 30	96.00 15	76.61 36	89.28 34	87.73 17	90.04 58	91.80 112	78.71 38	94.36 28	93.82 17	94.48 37	94.32 6

MP-MVS		90.84 6	91.95 34	89.55 3	92.92 4	90.90 19	96.56 6	79.60 11	86.83 58	88.75 12	89.00 73	94.38 77	84.01 9	94.94 24	94.34 10	95.45 24	93.24 23
Rongxuan Tan, Qing Wang, et al.: MP-MVS: Multi-Scale Windows PatchMatch and Planar Prior Multi-View Stereo.
SteuartSystems-ACMMP		90.00 17	91.73 35	87.97 12	91.21 29	90.29 28	96.51 7	78.00 23	86.33 61	85.32 40	88.23 82	94.67 69	82.08 20	95.13 22	93.88 16	94.72 35	93.59 13
Skip Steuart: Steuart Systems R&D Blog.
LS3D		89.02 33	91.69 36	85.91 30	89.72 43	90.81 20	92.56 44	71.69 66	90.83 21	87.24 22	89.71 63	92.07 106	78.37 42	94.43 27	92.59 27	95.86 13	91.35 41

PGM-MVS		90.42 11	91.58 37	89.05 5	91.77 14	91.06 13	96.51 7	78.94 16	85.41 70	87.67 18	87.02 96	95.26 56	83.62 12	95.01 23	93.94 15	95.79 19	93.40 20

CSCG		88.12 45	91.45 38	84.23 48	88.12 61	90.59 25	90.57 60	68.60 89	91.37 15	83.45 52	89.94 59	95.14 61	78.71 38	91.45 58	88.21 72	95.96 12	93.44 19

UA-Net		89.02 33	91.44 39	86.20 28	94.88 1	89.84 34	94.76 29	77.45 28	85.41 70	74.79 105	88.83 77	88.90 136	78.67 40	96.06 7	95.45 4	96.66 3	95.58 2

MSP-MVS		88.51 42	91.36 40	85.19 39	90.63 36	92.01 4	95.29 22	77.52 27	90.48 26	80.21 76	90.21 57	96.08 34	76.38 59	88.30 96	91.42 41	91.12 87	91.01 44
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
OMC-MVS		88.16 43	91.34 41	84.46 46	86.85 70	90.63 23	93.01 41	67.00 103	90.35 27	87.40 21	86.86 99	96.35 29	77.66 49	92.63 47	90.84 46	94.84 30	91.68 38

APD-MVS		89.14 29	91.25 42	86.67 24	91.73 15	91.02 15	95.50 20	77.74 24	84.04 82	79.47 82	91.48 45	94.85 66	81.14 25	92.94 41	92.20 35	94.47 38	92.24 32
Yuesong Wang, Zhaojie Zeng and etc.: Adaptive Patch Deformation for Textureless-Resilient Multi-View Stereo. CVPR2023
UniMVSNet_ETH3D		85.39 62	91.12 43	78.71 99	90.48 37	83.72 79	81.76 139	82.41 6	93.84 6	64.43 158	95.41 7	98.76 1	63.72 141	93.63 33	89.74 57	89.47 105	82.74 111

SF-MVS		87.85 48	90.95 44	84.22 49	88.17 60	87.90 53	90.80 56	71.80 65	89.28 34	82.70 57	89.90 60	95.37 54	77.91 47	91.69 54	90.04 54	93.95 44	92.47 29

X-MVS		89.36 28	90.73 45	87.77 16	91.50 20	91.23 8	96.76 4	78.88 17	87.29 53	87.14 25	78.98 146	94.53 71	76.47 57	95.25 19	94.28 11	95.85 14	93.55 16

anonymousdsp		85.62 59	90.53 46	79.88 92	64.64 207	76.35 142	96.28 12	53.53 192	85.63 67	81.59 69	92.81 30	97.71 12	86.88 2	94.56 25	92.83 24	96.35 6	93.84 9

CPTT-MVS		89.63 25	90.52 47	88.59 6	90.95 31	90.74 21	95.71 16	79.13 15	87.70 49	85.68 38	80.05 141	95.74 45	84.77 6	94.28 29	92.68 26	95.28 26	92.45 31

v7n		87.11 50	90.46 48	83.19 56	85.22 84	83.69 80	90.03 73	68.20 95	91.01 19	86.71 33	94.80 10	98.46 4	77.69 48	91.10 65	85.98 92	91.30 83	88.19 65

DeepPCF-MVS	81.61 6	87.95 47	90.29 49	85.22 38	87.48 65	90.01 31	93.79 34	73.54 54	88.93 38	83.89 45	89.40 67	90.84 121	80.26 31	90.62 72	90.19 53	92.36 70	92.03 35

3Dnovator+	83.71 3	88.13 44	90.00 50	85.94 29	86.82 71	91.06 13	94.26 33	75.39 46	88.85 40	85.76 37	85.74 109	86.92 145	78.02 45	93.03 40	92.21 34	95.39 25	92.21 34

DeepC-MVS_fast	81.78 5	87.38 49	89.64 51	84.75 41	89.89 42	90.70 22	92.74 43	74.45 50	86.02 64	82.16 64	86.05 106	91.99 110	75.84 65	91.16 63	90.44 49	93.41 51	91.09 43
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
HPM-MVS++		88.74 40	89.54 52	87.80 15	92.58 6	85.69 69	95.10 26	78.01 22	87.08 55	87.66 19	87.89 85	92.07 106	80.28 30	90.97 69	91.41 43	93.17 57	91.69 37

TranMVSNet+NR-MVSNet		85.23 65	89.38 53	80.39 90	88.78 53	83.77 78	87.40 95	76.75 34	85.47 68	68.99 140	95.18 8	97.55 16	67.13 124	91.61 56	89.13 65	93.26 54	82.95 108

Gipuma		86.47 55	89.25 54	83.23 55	83.88 102	78.78 123	85.35 115	68.42 91	92.69 10	89.03 11	91.94 38	96.32 32	81.80 21	94.45 26	86.86 82	90.91 88	83.69 99
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
MSLP-MVS++		86.29 57	89.10 55	83.01 58	85.71 81	89.79 35	87.04 103	74.39 51	85.17 72	78.92 86	77.59 155	93.57 85	82.60 17	93.23 36	91.88 39	89.42 106	92.46 30

CNVR-MVS		86.93 51	88.98 56	84.54 44	90.11 40	87.41 57	93.23 40	73.47 55	86.31 62	82.25 61	82.96 129	92.15 104	76.04 62	91.69 54	90.69 47	92.17 73	91.64 39

CNLPA		85.50 61	88.58 57	81.91 71	84.55 90	87.52 56	90.89 54	63.56 139	88.18 45	84.06 44	83.85 126	91.34 118	76.46 58	91.27 60	89.00 66	91.96 74	88.88 61

UniMVSNet (Re)		84.95 67	88.53 58	80.78 81	87.82 63	84.21 75	88.03 88	76.50 37	81.18 111	69.29 138	92.63 34	96.83 22	69.07 114	91.23 62	89.60 60	93.97 43	84.00 97

CDPH-MVS		86.66 54	88.52 59	84.48 45	89.61 45	88.27 46	92.86 42	72.69 61	80.55 118	82.71 56	86.92 98	93.32 89	75.55 67	91.00 68	89.85 56	93.47 49	89.71 53

ambc			88.38 60		91.62 17	87.97 52	84.48 122		88.64 43	87.93 15	87.38 91	94.82 68	74.53 76	89.14 88	83.86 115	85.94 150	86.84 75

TAPA-MVS	78.00 13	85.88 58	88.37 61	82.96 60	84.69 86	88.62 43	90.62 58	64.22 129	89.15 37	88.05 14	78.83 148	93.71 82	76.20 61	90.11 80	88.22 71	94.00 41	89.97 51
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
TSAR-MVS + COLMAP		85.51 60	88.36 62	82.19 67	86.05 78	87.69 54	90.50 65	70.60 72	86.40 60	82.33 59	89.69 64	92.52 98	74.01 81	87.53 100	86.84 83	89.63 101	87.80 71

pmmvs6		80.46 110	88.34 63	71.26 146	81.96 124	77.51 131	77.54 166	68.83 86	93.72 7	55.92 179	93.94 18	98.03 9	55.94 169	89.21 87	85.61 96	87.36 134	80.38 129

DU-MVS		84.88 68	88.27 64	80.92 79	88.30 57	83.59 81	87.06 101	78.35 19	80.64 116	70.49 132	92.67 32	96.91 21	68.13 117	91.79 51	89.29 64	93.20 55	83.02 105

PHI-MVS		86.37 56	88.14 65	84.30 47	86.65 73	87.56 55	90.76 57	70.16 73	82.55 92	89.65 7	84.89 117	92.40 99	75.97 63	90.88 70	89.70 58	92.58 65	89.03 60

Vis-MVSNet		83.32 84	88.12 66	77.71 106	77.91 157	83.44 83	90.58 59	69.49 78	81.11 112	67.10 152	89.85 61	91.48 116	71.71 98	91.34 59	89.37 62	89.48 104	90.26 49
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
UniMVSNet_NR-MVSNet		84.62 71	88.00 67	80.68 85	88.18 59	83.83 77	87.06 101	76.47 38	81.46 107	70.49 132	93.24 23	95.56 48	68.13 117	90.43 73	88.47 68	93.78 45	83.02 105

NCCC		86.74 52	87.97 68	85.31 36	90.64 35	87.25 58	93.27 39	74.59 49	86.50 59	83.72 46	75.92 172	92.39 100	77.08 53	91.72 53	90.68 48	92.57 67	91.30 42

train_agg		86.67 53	87.73 69	85.43 35	91.51 19	82.72 87	94.47 31	74.22 53	81.71 100	81.54 70	89.20 71	92.87 94	78.33 43	90.12 79	88.47 68	92.51 69	89.04 59

EG-PatchMatch MVS		84.35 72	87.55 70	80.62 86	86.38 75	82.24 92	86.75 104	64.02 134	84.24 78	78.17 91	89.38 68	95.03 64	78.78 37	89.95 81	86.33 89	89.59 102	85.65 84

PLC	76.06 15	85.38 63	87.46 71	82.95 61	85.79 80	88.84 41	88.86 83	68.70 88	87.06 56	83.60 48	79.02 144	90.05 127	77.37 52	90.88 70	89.66 59	93.37 52	86.74 76
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
NR-MVSNet		82.89 89	87.43 72	77.59 108	83.91 101	83.59 81	87.10 100	78.35 19	80.64 116	68.85 141	92.67 32	96.50 24	54.19 179	87.19 106	88.68 67	93.16 58	82.75 110

TSAR-MVS + GP.		85.32 64	87.41 73	82.89 62	90.07 41	85.69 69	89.07 81	72.99 60	82.45 93	74.52 108	85.09 114	87.67 142	79.24 33	91.11 64	90.41 50	91.45 79	89.45 55

CLD-MVS		82.75 93	87.22 74	77.54 109	88.01 62	85.76 68	90.23 69	54.52 186	82.28 96	82.11 65	88.48 80	95.27 55	63.95 139	89.41 85	88.29 70	86.45 143	81.01 125
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
TransMVSNet (Re)		79.05 126	86.66 75	70.18 156	83.32 109	75.99 145	77.54 166	63.98 135	90.68 24	55.84 180	94.80 10	96.06 35	53.73 182	86.27 113	83.22 123	86.65 138	79.61 139

Baseline_NR-MVSNet		82.79 91	86.51 76	78.44 103	88.30 57	75.62 150	87.81 90	74.97 48	81.53 104	66.84 153	94.71 12	96.46 25	66.90 125	91.79 51	83.37 122	85.83 152	82.09 116

MCST-MVS		84.79 69	86.48 77	82.83 63	87.30 67	87.03 61	90.46 67	69.33 81	83.14 86	82.21 63	81.69 137	92.14 105	75.09 72	87.27 103	84.78 106	92.58 65	89.30 57

EPP-MVSNet		82.76 92	86.47 78	78.45 102	86.00 79	84.47 74	85.39 114	68.42 91	84.17 79	62.97 162	89.26 70	76.84 182	72.13 94	92.56 48	90.40 51	95.76 20	87.56 73

HQP-MVS		85.02 66	86.41 79	83.40 54	89.19 48	86.59 63	91.28 50	71.60 67	82.79 89	83.48 51	78.65 150	93.54 86	72.55 89	86.49 111	85.89 95	92.28 72	90.95 46

FC-MVSNet-train		79.20 125	86.29 80	70.94 150	84.06 96	77.67 130	85.68 109	64.11 131	82.90 88	52.22 194	92.57 35	93.69 83	49.52 194	88.30 96	86.93 80	90.03 95	81.95 118

MVS_0304		84.73 70	86.19 81	83.02 57	88.32 56	86.71 62	91.55 48	70.87 70	73.79 147	82.88 54	85.13 113	93.35 88	72.55 89	88.62 91	87.69 74	91.93 75	88.05 69

TinyColmap		83.79 76	86.12 82	81.07 78	83.42 108	81.44 98	85.42 113	68.55 90	88.71 42	89.46 8	87.60 87	92.72 95	70.34 108	89.29 86	81.94 131	89.20 107	81.12 124

MVS_111021_HR		83.95 75	86.10 83	81.44 76	84.62 88	80.29 110	90.51 64	68.05 96	84.07 81	80.38 74	84.74 120	91.37 117	74.23 77	90.37 75	87.25 78	90.86 89	84.59 89

3Dnovator	79.41 10	82.21 95	86.07 84	77.71 106	79.31 142	84.61 73	87.18 98	61.02 160	85.65 66	76.11 96	85.07 115	85.38 154	70.96 104	87.22 104	86.47 85	91.66 77	88.12 68

canonicalmvs		81.22 108	86.04 85	75.60 119	83.17 112	83.18 84	80.29 149	65.82 116	85.97 65	67.98 148	77.74 154	91.51 115	65.17 135	88.62 91	86.15 91	91.17 86	89.09 58

UGNet		79.62 119	85.91 86	72.28 143	73.52 178	83.91 76	86.64 105	69.51 77	79.85 123	62.57 164	85.82 108	89.63 128	53.18 183	88.39 95	87.35 77	88.28 125	86.43 78
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
Anonymous20231211		79.37 121	85.78 87	71.89 144	82.87 117	79.66 116	78.77 163	63.93 137	83.36 84	59.39 169	90.54 53	94.66 70	56.46 167	87.38 101	84.12 111	89.92 97	80.74 126

casdiffmvs_mvg		81.50 103	85.70 88	76.60 115	82.68 118	80.54 107	83.50 126	64.49 127	83.40 83	72.53 118	92.15 37	95.40 52	65.84 132	84.69 130	81.89 132	90.59 90	81.86 120
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
pm-mvs1		78.21 131	85.68 89	69.50 161	80.38 133	75.73 148	76.25 174	65.04 121	87.59 50	54.47 185	93.16 25	95.99 40	54.20 178	86.37 112	82.98 126	86.64 139	77.96 148

DCV-MVSNet		80.04 113	85.67 90	73.48 137	82.91 115	81.11 103	80.44 148	66.06 110	85.01 73	62.53 165	78.84 147	94.43 76	58.51 160	88.66 90	85.91 93	90.41 91	85.73 83

MVS_111021_LR		83.20 86	85.33 91	80.73 84	82.88 116	78.23 127	89.61 75	65.23 120	82.08 97	81.19 71	85.31 111	92.04 109	75.22 69	89.50 83	85.90 94	90.24 92	84.23 93

PCF-MVS	76.59 14	84.11 74	85.27 92	82.76 64	86.12 77	88.30 45	91.24 51	69.10 82	82.36 95	84.45 43	77.56 156	90.40 126	72.91 88	85.88 116	83.88 113	92.72 64	88.53 63
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
v1192		83.61 78	85.23 93	81.72 73	84.05 97	82.15 93	89.54 76	66.20 108	81.38 109	86.76 32	91.79 42	96.03 36	74.88 74	81.81 151	80.92 139	88.91 113	82.50 113

v10		83.17 87	85.22 94	80.78 81	83.26 110	82.99 85	88.66 85	66.49 106	79.24 127	83.60 48	91.46 46	95.47 50	74.12 78	82.60 146	80.66 140	88.53 122	84.11 96

AdaColmap		84.15 73	85.14 95	83.00 59	89.08 49	87.14 60	90.56 61	70.90 69	82.40 94	80.41 73	73.82 183	84.69 156	75.19 70	91.58 57	89.90 55	91.87 76	86.48 77

v1144		83.22 85	85.01 96	81.14 77	83.76 105	81.60 96	88.95 82	65.58 118	81.89 99	85.80 36	91.68 44	95.84 41	74.04 80	82.12 148	80.56 142	88.70 117	81.41 122

IS_MVSNet		81.72 101	85.01 96	77.90 105	86.19 76	82.64 89	85.56 110	70.02 74	80.11 121	63.52 160	87.28 93	81.18 167	67.26 122	91.08 67	89.33 63	94.82 31	83.42 102

v144192		83.43 83	84.97 98	81.63 75	83.43 107	81.23 101	89.42 79	66.04 112	81.45 108	86.40 34	91.46 46	95.70 46	75.76 66	82.14 147	80.23 146	88.74 115	82.57 112

v1921920		83.49 82	84.94 99	81.80 72	83.78 104	81.20 102	89.50 77	65.91 113	81.64 102	87.18 24	91.70 43	95.39 53	75.85 64	81.56 154	80.27 145	88.60 118	82.80 109

v1240		83.57 80	84.94 99	81.97 70	84.05 97	81.27 100	89.46 78	66.06 110	81.31 110	87.50 20	91.88 41	95.46 51	76.25 60	81.16 156	80.51 143	88.52 123	82.98 107

CS-MVS-test		83.59 79	84.86 101	82.10 69	83.04 113	81.05 104	91.58 47	67.48 102	72.52 154	78.42 89	84.75 119	91.82 111	78.62 41	91.98 50	87.54 76	93.48 48	84.35 92

FMVSNet1		78.20 132	84.83 102	70.46 154	78.62 149	79.03 120	77.90 165	67.53 101	83.02 87	55.10 183	87.19 95	93.18 91	55.65 172	85.57 117	83.39 119	87.98 127	82.40 114

CS-MVS		83.57 80	84.79 103	82.14 68	83.83 103	81.48 97	87.29 96	66.54 105	72.73 153	80.05 78	84.04 124	93.12 93	80.35 28	89.50 83	86.34 88	94.76 34	86.32 80

DROMVSNet		83.70 77	84.77 104	82.46 66	87.47 66	82.79 86	85.50 111	72.00 63	69.81 165	77.66 92	85.02 116	89.63 128	78.14 44	90.40 74	87.56 75	94.00 41	88.16 66

Anonymous202405211			84.68 105		83.92 100	79.45 118	79.03 161	67.79 98	82.01 98		88.77 79	92.58 97	55.93 170	86.68 109	84.26 110	88.92 112	78.98 141

CANet		82.84 90	84.60 106	80.78 81	87.30 67	85.20 72	90.23 69	69.00 83	72.16 157	78.73 87	84.49 122	90.70 124	69.54 112	87.65 99	86.17 90	89.87 98	85.84 82

v8		82.20 96	84.56 107	79.45 95	82.42 120	81.65 95	87.26 97	64.27 128	79.36 126	81.70 68	91.04 52	95.75 44	73.30 87	82.82 142	79.18 153	87.74 130	82.09 116

test1111		79.67 117	84.40 108	74.16 132	85.29 83	79.56 117	81.16 143	73.13 59	84.65 77	56.08 177	88.38 81	86.14 149	60.49 151	89.78 82	85.59 97	88.79 114	76.68 151

QAPM		80.43 111	84.34 109	75.86 117	79.40 141	82.06 94	79.86 154	61.94 154	83.28 85	74.73 107	81.74 136	85.44 153	70.97 103	84.99 128	84.71 108	88.29 124	88.14 67

thisisatest0515		81.18 109	84.32 110	77.52 110	76.73 168	74.84 157	85.06 118	61.37 157	81.05 113	73.95 110	88.79 78	89.25 133	75.49 68	85.98 115	84.78 106	92.53 68	85.56 85

tfpnnormal		77.16 135	84.26 111	68.88 164	81.02 130	75.02 154	76.52 173	63.30 142	87.29 53	52.40 192	91.24 50	93.97 79	54.85 176	85.46 120	81.08 137	85.18 158	75.76 157

v2v482		82.20 96	84.26 111	79.81 93	82.67 119	80.18 111	87.67 92	63.96 136	81.69 101	84.73 41	91.27 49	96.33 31	72.05 95	81.94 150	79.56 150	87.79 129	78.84 142

MSDG		81.39 106	84.23 113	78.09 104	82.40 121	82.47 91	85.31 117	60.91 161	79.73 124	80.26 75	86.30 102	88.27 140	69.67 110	87.20 105	84.98 103	89.97 96	80.67 127

ECVR-MVS		79.31 124	84.20 114	73.60 134	84.55 90	80.37 108	79.63 157	73.23 57	82.64 90	55.98 178	87.50 88	86.85 146	59.61 155	90.35 76	86.46 86	88.58 120	75.26 160

PVSNet_Blended_VisFu		83.00 88	84.16 115	81.65 74	82.17 123	86.01 66	88.03 88	71.23 68	76.05 140	79.54 81	83.88 125	83.44 157	77.49 51	87.38 101	84.93 104	91.41 80	87.40 74

casdiffmvs		79.93 114	84.11 116	75.05 124	81.41 129	78.99 121	82.95 131	62.90 147	81.53 104	68.60 145	91.94 38	96.03 36	65.84 132	82.89 141	77.07 164	88.59 119	80.34 133
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
FPMVS		81.56 102	84.04 117	78.66 100	82.92 114	75.96 146	86.48 107	65.66 117	84.67 76	71.47 127	77.78 153	83.22 160	77.57 50	91.24 61	90.21 52	87.84 128	85.21 86

GeoE		81.92 100	83.87 118	79.66 94	84.64 87	79.87 112	89.75 74	65.90 114	76.12 139	75.87 98	84.62 121	92.23 102	71.96 96	86.83 108	83.60 116	89.83 99	83.81 98

Effi-MVS+		82.33 94	83.87 118	80.52 88	84.51 93	81.32 99	87.53 93	68.05 96	74.94 145	79.67 80	82.37 134	92.31 101	72.21 91	85.06 123	86.91 81	91.18 85	84.20 94

Fast-Effi-MVS+		81.42 104	83.82 120	78.62 101	82.24 122	80.62 106	87.72 91	63.51 140	73.01 149	74.75 106	83.80 127	92.70 96	73.44 86	88.15 98	85.26 100	90.05 94	83.17 103

DELS-MVS		79.71 116	83.74 121	75.01 126	79.31 142	82.68 88	84.79 120	60.06 167	75.43 143	69.09 139	86.13 104	89.38 131	67.16 123	85.12 122	83.87 114	89.65 100	83.57 100
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
PM-MVS		80.42 112	83.63 122	76.67 113	78.04 154	72.37 167	87.14 99	60.18 166	80.13 120	71.75 125	86.12 105	93.92 81	77.08 53	86.56 110	85.12 102	85.83 152	81.18 123

FC-MVSNet-test		75.91 146	83.59 123	66.95 175	76.63 170	69.07 177	85.33 116	64.97 122	84.87 75	41.95 208	93.17 24	87.04 144	47.78 197	91.09 66	85.56 98	85.06 159	74.34 161

V42		79.59 120	83.59 123	74.93 129	69.61 191	77.05 138	86.59 106	55.84 181	78.42 131	77.29 93	89.84 62	95.08 62	74.12 78	83.05 139	80.11 148	86.12 146	81.59 121

Effi-MVS+-dtu		82.04 98	83.39 125	80.48 89	85.48 82	86.57 64	88.40 86	68.28 93	69.04 172	73.13 117	76.26 167	91.11 120	74.74 75	88.40 94	87.76 73	92.84 63	84.57 90

USDC		81.39 106	83.07 126	79.43 96	81.48 127	78.95 122	82.62 134	66.17 109	87.45 52	90.73 4	82.40 133	93.65 84	66.57 127	83.63 138	77.97 156	89.00 111	77.45 150

MAR-MVS		81.98 99	82.92 127	80.88 80	85.18 85	85.85 67	89.13 80	69.52 76	71.21 161	82.25 61	71.28 193	88.89 137	69.69 109	88.71 89	86.96 79	89.52 103	87.57 72
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
MDA-MVSNet-bldmvs		76.51 139	82.87 128	69.09 163	50.71 218	74.72 159	84.05 124	60.27 165	81.62 103	71.16 129	88.21 83	91.58 113	69.62 111	92.78 44	77.48 161	78.75 180	73.69 166

IterMVS-LS		79.79 115	82.56 129	76.56 116	81.83 125	77.85 129	79.90 153	69.42 80	78.93 129	71.21 128	90.47 54	85.20 155	70.86 105	80.54 161	80.57 141	86.15 145	84.36 91
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
v148		79.33 123	82.32 130	75.84 118	80.14 134	75.74 147	81.98 138	57.06 178	81.51 106	79.36 83	89.42 66	96.42 27	71.32 99	81.54 155	75.29 173	85.20 157	76.32 152

DPM-MVS		81.42 104	82.11 131	80.62 86	87.54 64	85.30 71	90.18 71	68.96 84	81.00 114	79.15 84	70.45 199	83.29 159	67.67 121	82.81 143	83.46 117	90.19 93	88.48 64

pmmvs-eth3d		79.64 118	82.06 132	76.83 112	80.05 135	72.64 165	87.47 94	66.59 104	80.83 115	73.50 113	89.32 69	93.20 90	67.78 119	80.78 159	81.64 135	85.58 155	76.01 153

MIMVSNet1		73.40 158	81.85 133	63.55 187	72.90 181	64.37 191	84.58 121	53.60 191	90.84 20	53.92 186	87.75 86	96.10 33	45.31 200	85.37 121	79.32 152	70.98 195	69.18 180

diffmvs		76.74 137	81.61 134	71.06 148	75.64 173	74.45 160	80.68 147	57.57 177	77.48 132	67.62 151	88.95 74	93.94 80	61.98 148	79.74 164	76.18 168	82.85 170	80.50 128
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
OpenMVS	75.38 16	78.44 130	81.39 135	74.99 127	80.46 132	79.85 113	79.99 151	58.31 175	77.34 134	73.85 111	77.19 159	82.33 165	68.60 116	84.67 131	81.95 130	88.72 116	86.40 79

Vis-MVSNet (Re-imp)		76.15 143	80.84 136	70.68 151	83.66 106	74.80 158	81.66 141	69.59 75	80.48 119	46.94 203	87.44 90	80.63 169	53.14 184	86.87 107	84.56 109	89.12 108	71.12 171

FA-MVS(training)		78.93 128	80.63 137	76.93 111	79.79 138	75.57 151	85.44 112	61.95 153	77.19 135	78.97 85	84.82 118	82.47 162	66.43 130	84.09 135	80.13 147	89.02 110	80.15 136

EU-MVSNet		76.48 140	80.53 138	71.75 145	67.62 197	70.30 172	81.74 140	54.06 189	75.47 142	71.01 130	80.10 139	93.17 92	73.67 83	83.73 137	77.85 157	82.40 171	83.07 104

FMVSNet2		74.43 155	79.70 139	68.27 167	76.76 162	77.36 133	75.77 178	65.36 119	72.28 155	52.97 189	81.92 135	85.61 152	52.73 187	80.66 160	79.73 149	86.04 147	80.37 130

DI_MVS_plusplus_trai		77.64 133	79.64 140	75.31 122	79.87 137	76.89 139	81.55 142	63.64 138	76.21 138	72.03 123	85.59 110	82.97 161	66.63 126	79.27 167	77.78 158	88.14 126	78.76 144

EPNet		79.36 122	79.44 141	79.27 98	89.51 46	77.20 136	88.35 87	77.35 31	68.27 174	74.29 109	76.31 165	79.22 172	59.63 154	85.02 127	85.45 99	86.49 142	84.61 88
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
MVS_Test		76.72 138	79.40 142	73.60 134	78.85 148	74.99 155	79.91 152	61.56 156	69.67 166	72.44 119	85.98 107	90.78 122	63.50 144	78.30 169	75.74 171	85.33 156	80.31 134

IterMVS-SCA-FT		77.23 134	79.18 143	74.96 128	76.67 169	79.85 113	75.58 183	61.34 158	73.10 148	73.79 112	86.23 103	79.61 171	79.00 36	80.28 163	75.50 172	83.41 169	79.70 138

CANet_DTU		75.04 152	78.45 144	71.07 147	77.27 159	77.96 128	83.88 125	58.00 176	64.11 191	68.67 144	75.65 174	88.37 139	53.92 181	82.05 149	81.11 136	84.67 160	79.88 137

thres600view7		74.34 156	78.43 145	69.56 160	80.47 131	76.28 143	78.65 164	62.56 149	77.39 133	52.53 190	74.03 181	76.78 183	55.90 171	85.06 123	85.19 101	87.25 135	74.29 162

PVSNet_BlendedMVS		76.45 141	78.12 146	74.49 130	76.76 162	78.46 124	79.65 155	63.26 143	65.42 186	73.15 115	75.05 177	88.96 134	66.51 128	82.73 144	77.66 159	87.61 131	78.60 145

PVSNet_Blended		76.45 141	78.12 146	74.49 130	76.76 162	78.46 124	79.65 155	63.26 143	65.42 186	73.15 115	75.05 177	88.96 134	66.51 128	82.73 144	77.66 159	87.61 131	78.60 145

ETV-MVS		79.01 127	77.98 148	80.22 91	86.69 72	79.73 115	88.80 84	68.27 94	63.22 195	71.56 126	70.25 201	73.63 192	73.66 84	90.30 78	86.77 84	92.33 71	81.95 118

EIA-MVS		78.57 129	77.90 149	79.35 97	87.24 69	80.71 105	86.16 108	64.03 133	62.63 200	73.49 114	73.60 184	76.12 186	73.83 82	88.49 93	84.93 104	91.36 81	78.78 143

GBi-Net		73.17 160	77.64 150	67.95 170	76.76 162	77.36 133	75.77 178	64.57 124	62.99 197	51.83 195	76.05 168	77.76 178	52.73 187	85.57 117	83.39 119	86.04 147	80.37 130

test1		73.17 160	77.64 150	67.95 170	76.76 162	77.36 133	75.77 178	64.57 124	62.99 197	51.83 195	76.05 168	77.76 178	52.73 187	85.57 117	83.39 119	86.04 147	80.37 130

CVMVSNet		75.65 148	77.62 152	73.35 140	71.95 184	69.89 174	83.04 130	60.84 162	69.12 170	68.76 142	79.92 142	78.93 174	73.64 85	81.02 157	81.01 138	81.86 174	83.43 101

pmmvs4		75.92 145	77.48 153	74.10 133	78.21 153	70.94 169	84.06 123	64.78 123	75.13 144	68.47 146	84.12 123	83.32 158	64.74 138	75.93 181	79.14 154	84.31 162	73.77 165

Fast-Effi-MVS+-dtu		76.92 136	77.18 154	76.62 114	79.55 139	79.17 119	84.80 119	77.40 29	64.46 190	68.75 143	70.81 197	86.57 147	63.36 146	81.74 152	81.76 133	85.86 151	75.78 156

CDS-MVSNet		73.07 163	77.02 155	68.46 166	81.62 126	72.89 164	79.56 159	70.78 71	69.56 167	52.52 191	77.37 158	81.12 168	42.60 202	84.20 134	83.93 112	83.65 165	70.07 176
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
IB-MVS	71.28 17	75.21 151	77.00 156	73.12 141	76.76 162	77.45 132	83.05 129	58.92 172	63.01 196	64.31 159	59.99 214	87.57 143	68.64 115	86.26 114	82.34 129	87.05 137	82.36 115
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
thres400		73.13 162	76.99 157	68.62 165	79.46 140	74.93 156	77.23 168	61.23 159	75.54 141	52.31 193	72.20 188	77.10 181	54.89 174	82.92 140	82.62 128	86.57 141	73.66 167

test2506		75.32 150	76.87 158	73.50 136	84.55 90	80.37 108	79.63 157	73.23 57	82.64 90	55.41 181	76.87 162	45.42 224	59.61 155	90.35 76	86.46 86	88.58 120	75.98 154

PatchMatch-RL		76.05 144	76.64 159	75.36 121	77.84 158	69.87 175	81.09 145	63.43 141	71.66 159	68.34 147	71.70 189	81.76 166	74.98 73	84.83 129	83.44 118	86.45 143	73.22 168

IterMVS		73.62 157	76.53 160	70.23 155	71.83 185	77.18 137	80.69 146	53.22 193	72.23 156	66.62 154	85.21 112	78.96 173	69.54 112	76.28 180	71.63 183	79.45 177	74.25 163
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
GA-MVS		75.01 153	76.39 161	73.39 138	78.37 150	75.66 149	80.03 150	58.40 174	70.51 163	75.85 99	83.24 128	76.14 185	63.75 140	77.28 173	76.62 167	83.97 164	75.30 159

CMPMVS	55.74 18	71.56 169	76.26 162	66.08 180	68.11 195	63.91 193	63.17 209	50.52 201	68.79 173	75.49 100	70.78 198	85.67 151	63.54 143	81.58 153	77.20 163	75.63 182	85.86 81
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
tttt0517		75.86 147	76.23 163	75.42 120	75.55 174	74.06 161	82.73 132	60.31 163	69.24 168	70.24 134	79.18 143	58.79 210	72.17 92	84.49 132	83.08 124	91.54 78	84.80 87

test20.03		69.91 173	76.20 164	62.58 188	84.01 99	67.34 183	75.67 182	65.88 115	79.98 122	40.28 212	82.65 130	89.31 132	39.63 207	77.41 172	73.28 177	69.98 196	63.40 191

testgi		68.20 182	76.05 165	59.04 194	79.99 136	67.32 184	81.16 143	51.78 197	84.91 74	39.36 213	73.42 185	95.19 57	32.79 213	76.54 178	70.40 186	69.14 199	64.55 187

thres200		72.41 166	76.00 166	68.21 168	78.28 151	76.28 143	74.94 184	62.56 149	72.14 158	51.35 198	69.59 203	76.51 184	54.89 174	85.06 123	80.51 143	87.25 135	71.92 170

thisisatest0530		75.54 149	75.95 167	75.05 124	75.08 175	73.56 162	82.15 137	60.31 163	69.17 169	69.32 137	79.02 144	58.78 211	72.17 92	83.88 136	83.08 124	91.30 83	84.20 94

MDTV_nov1_ep13_2view		72.96 164	75.59 168	69.88 157	71.15 188	64.86 190	82.31 136	54.45 187	76.30 137	78.32 90	86.52 100	91.58 113	61.35 149	76.80 174	66.83 194	71.70 188	66.26 184

tfpn200view9		72.01 167	75.40 169	68.06 169	77.97 155	76.44 141	77.04 170	62.67 148	66.81 177	50.82 199	67.30 205	75.67 188	52.46 190	85.06 123	82.64 127	87.41 133	73.86 164

baseline		69.33 177	75.37 170	62.28 190	66.54 203	66.67 186	73.95 187	48.07 202	66.10 180	59.26 170	82.45 131	86.30 148	54.44 177	74.42 184	73.25 178	71.42 191	78.43 147

gg-mvs-nofinetune		72.68 165	75.21 171	69.73 158	81.48 127	69.04 178	70.48 195	76.67 35	86.92 57	67.80 150	88.06 84	64.67 200	42.12 204	77.60 171	73.65 176	79.81 176	66.57 183

FMVSNet3		71.40 171	75.20 172	66.97 174	75.00 176	76.59 140	74.29 185	64.57 124	62.99 197	51.83 195	76.05 168	77.76 178	51.49 192	76.58 177	77.03 165	84.62 161	79.43 140

pmmvs5		68.91 178	74.35 173	62.56 189	67.45 199	66.78 185	71.70 191	51.47 198	67.17 176	56.25 176	82.41 132	88.59 138	47.21 199	73.21 191	74.23 174	81.30 175	68.03 182

ET-MVSNet_ETH3D		74.71 154	74.19 174	75.31 122	79.22 144	75.29 152	82.70 133	64.05 132	65.45 185	70.96 131	77.15 160	57.70 212	65.89 131	84.40 133	81.65 134	89.03 109	77.67 149

HyFIR lowres test		73.29 159	74.14 175	72.30 142	73.08 180	78.33 126	83.12 128	62.41 151	63.81 192	62.13 166	76.67 164	78.50 175	71.09 101	74.13 185	77.47 162	81.98 173	70.10 175

MS-PatchMatch		71.18 172	73.99 176	67.89 172	77.16 160	71.76 168	77.18 169	56.38 180	67.35 175	55.04 184	74.63 179	75.70 187	62.38 147	76.62 176	75.97 170	79.22 178	75.90 155

new-patchmatchnet		62.59 197	73.79 177	49.53 210	76.98 161	53.57 206	53.46 218	54.64 185	85.43 69	28.81 217	91.94 38	96.41 28	25.28 215	76.80 174	53.66 214	57.99 211	58.69 204

baseline1		69.62 175	73.55 178	65.02 186	78.95 147	70.39 171	71.38 194	62.03 152	70.97 162	47.95 202	78.47 151	68.19 198	47.77 198	79.65 166	76.94 166	82.05 172	70.27 174

Anonymous20231206		67.28 184	73.41 179	60.12 193	76.45 171	63.61 194	74.21 186	56.52 179	76.35 136	42.23 207	75.81 173	90.47 125	41.51 205	74.52 182	69.97 188	69.83 197	63.17 192

EPNet_dtu		71.90 168	73.03 180	70.59 152	78.28 151	61.64 196	82.44 135	64.12 130	63.26 194	69.74 135	71.47 191	82.41 163	51.89 191	78.83 168	78.01 155	77.07 181	75.60 158
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
thres100view900		69.86 174	72.97 181	66.24 177	77.97 155	72.49 166	73.29 188	59.12 170	66.81 177	50.82 199	67.30 205	75.67 188	50.54 193	78.24 170	79.40 151	85.71 154	70.88 172

CHOSEN 1792x2688		68.80 179	71.09 182	66.13 179	69.11 193	68.89 179	78.98 162	54.68 184	61.63 202	56.69 174	71.56 190	78.39 176	67.69 120	72.13 192	72.01 182	69.63 198	73.02 169

pmnet_mix02		62.60 196	70.81 183	53.02 206	66.56 202	50.44 212	62.81 210	46.84 204	79.13 128	43.76 206	87.45 89	90.75 123	39.85 206	70.48 197	57.09 208	58.27 210	60.32 201

gm-plane-assit		71.56 169	69.99 184	73.39 138	84.43 94	73.21 163	90.42 68	51.36 199	84.08 80	76.00 97	91.30 48	37.09 225	59.01 158	73.65 188	70.24 187	79.09 179	60.37 200

MVSTER		68.08 183	69.73 185	66.16 178	66.33 205	70.06 173	75.71 181	52.36 195	55.18 214	58.64 171	70.23 202	56.72 215	57.34 164	79.68 165	76.03 169	86.61 140	80.20 135

TAMVS		63.02 192	69.30 186	55.70 201	70.12 189	56.89 202	69.63 199	45.13 205	70.23 164	38.00 214	77.79 152	75.15 190	42.60 202	74.48 183	72.81 181	68.70 200	57.75 207

MIMVSNet		63.02 192	69.02 187	56.01 199	68.20 194	59.26 199	70.01 198	53.79 190	71.56 160	41.26 211	71.38 192	82.38 164	36.38 209	71.43 195	67.32 193	66.45 204	59.83 202

pmmvs3		62.72 195	68.71 188	55.74 200	50.74 217	57.10 201	70.05 197	28.82 215	61.57 204	57.39 173	71.19 195	85.73 150	53.96 180	73.36 190	69.43 190	73.47 186	62.55 194

baseline2		68.71 180	68.34 189	69.14 162	75.69 172	69.70 176	76.60 172	55.53 183	60.13 205	62.07 167	66.76 207	60.35 205	60.77 150	76.53 179	74.03 175	84.19 163	70.88 172

CR-MVSNet		69.56 176	68.34 189	70.99 149	72.78 183	67.63 181	64.47 207	67.74 99	59.93 206	72.30 120	80.10 139	56.77 214	65.04 136	71.64 193	72.91 179	83.61 167	69.40 178

SCA		68.54 181	67.52 191	69.73 158	67.79 196	75.04 153	76.96 171	68.94 85	66.41 179	67.86 149	74.03 181	60.96 203	65.55 134	68.99 201	65.67 195	71.30 193	61.54 199

CostFormer		66.81 186	66.94 192	66.67 176	72.79 182	68.25 180	79.55 160	55.57 182	65.52 184	62.77 163	76.98 161	60.09 206	56.73 166	65.69 209	62.35 198	72.59 187	69.71 177

PatchT		66.25 187	66.76 193	65.67 183	55.87 213	60.75 197	70.17 196	59.00 171	59.80 208	72.30 120	78.68 149	54.12 219	65.04 136	71.64 193	72.91 179	71.63 190	69.40 178

N_pmnet		54.95 211	65.90 194	42.18 211	66.37 204	43.86 218	57.92 215	39.79 210	79.54 125	17.24 222	86.31 101	87.91 141	25.44 214	64.68 210	51.76 216	46.33 217	47.23 214

PMMVS		61.98 199	65.61 195	57.74 196	45.03 219	51.76 210	69.54 200	35.05 212	55.49 213	55.32 182	68.23 204	78.39 176	58.09 161	70.21 199	71.56 184	83.42 168	63.66 189

test0.0.03 1		61.79 200	65.33 196	57.65 197	79.07 145	64.09 192	68.51 204	62.93 145	61.59 203	33.71 216	61.58 213	71.58 196	33.43 212	70.95 196	68.68 191	68.26 201	58.82 203

MDTV_nov1_ep13		64.96 189	64.77 197	65.18 185	67.08 200	62.46 195	75.80 177	51.10 200	62.27 201	69.74 135	74.12 180	62.65 201	55.64 173	68.19 203	62.16 202	71.70 188	61.57 198

dps		65.14 188	64.50 198	65.89 182	71.41 187	65.81 189	71.44 193	61.59 155	58.56 209	61.43 168	75.45 175	52.70 221	58.06 162	69.57 200	64.65 196	71.39 192	64.77 186

PMMVS2		48.13 214	64.06 199	29.55 214	44.06 220	36.69 220	51.95 219	29.97 214	74.75 146	8.90 224	76.02 171	91.24 119	7.53 218	73.78 187	55.91 209	34.87 219	40.01 218

RPMNet		67.02 185	63.99 200	70.56 153	71.55 186	67.63 181	75.81 176	69.44 79	59.93 206	63.24 161	64.32 209	47.51 223	59.68 153	70.37 198	69.64 189	83.64 166	68.49 181

PatchmatchNet		64.81 190	63.74 201	66.06 181	69.21 192	58.62 200	73.16 189	60.01 168	65.92 181	66.19 156	76.27 166	59.09 207	60.45 152	66.58 206	61.47 204	67.33 202	58.24 205
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
tpm		62.79 194	63.25 202	62.26 191	70.09 190	53.78 205	71.65 192	47.31 203	65.72 183	76.70 94	80.62 138	56.40 217	48.11 196	64.20 211	58.54 205	59.70 208	63.47 190

new_pmnet		52.29 212	63.16 203	39.61 213	58.89 211	44.70 217	48.78 220	34.73 213	65.88 182	17.85 221	73.42 185	80.00 170	23.06 216	67.00 205	62.28 201	54.36 213	48.81 213

E-PMN		59.07 204	62.79 204	54.72 202	67.01 201	47.81 215	60.44 213	43.40 206	72.95 150	44.63 205	70.42 200	73.17 193	58.73 159	80.97 158	51.98 215	54.14 214	42.26 216

tpm cat1		64.79 191	62.74 205	67.17 173	74.61 177	65.91 188	76.18 175	59.32 169	64.88 189	66.41 155	71.21 194	53.56 220	59.17 157	61.53 213	58.16 207	67.33 202	63.95 188

EMVS		58.97 205	62.63 206	54.70 203	66.26 206	48.71 213	61.74 211	42.71 207	72.80 152	46.00 204	73.01 187	71.66 194	57.91 163	80.41 162	50.68 217	53.55 215	41.11 217

test-mter		59.39 203	61.59 207	56.82 198	53.21 214	54.82 204	73.12 190	26.57 217	53.19 215	56.31 175	64.71 208	60.47 204	56.36 168	68.69 202	64.27 197	75.38 183	65.00 185

ADS-MVSNet		56.89 207	61.09 208	52.00 208	59.48 210	48.10 214	58.02 214	54.37 188	72.82 151	49.19 201	75.32 176	65.97 199	37.96 208	59.34 216	54.66 212	52.99 216	51.42 212

MVE	41.12 19	51.80 213	60.92 209	41.16 212	35.21 221	34.14 221	48.45 221	41.39 209	69.11 171	19.53 220	63.33 210	73.80 191	63.56 142	67.19 204	61.51 203	38.85 218	57.38 208
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
GG-mvs-BLEND		41.63 215	60.36 210	19.78 215	0.14 226	66.04 187	55.66 217	0.17 223	57.64 210	2.42 225	51.82 215	69.42 197	0.28 222	64.11 212	58.29 206	60.02 207	55.18 209

FMVSNet5		56.37 209	60.14 211	51.98 209	60.83 209	59.58 198	66.85 206	42.37 208	52.68 216	41.33 210	47.09 217	54.68 218	35.28 210	73.88 186	70.77 185	65.24 205	62.26 195

tpmrst		59.42 202	60.02 212	58.71 195	67.56 198	53.10 207	66.99 205	51.88 196	63.80 193	57.68 172	76.73 163	56.49 216	48.73 195	56.47 217	55.55 210	59.43 209	58.02 206

EPMVS		56.62 208	59.77 213	52.94 207	62.41 208	50.55 211	60.66 212	52.83 194	65.15 188	41.80 209	77.46 157	57.28 213	42.68 201	59.81 215	54.82 211	57.23 212	53.35 210

test-LLR		62.15 198	59.46 214	65.29 184	79.07 145	52.66 208	69.46 201	62.93 145	50.76 217	53.81 187	63.11 211	58.91 208	52.87 185	66.54 207	62.34 199	73.59 184	61.87 196

TESTMET0.1,1		57.21 206	59.46 214	54.60 204	50.95 216	52.66 208	69.46 201	26.91 216	50.76 217	53.81 187	63.11 211	58.91 208	52.87 185	66.54 207	62.34 199	73.59 184	61.87 196

CHOSEN 280x420		56.32 210	58.85 216	53.36 205	51.63 215	39.91 219	69.12 203	38.61 211	56.29 211	36.79 215	48.84 216	62.59 202	63.39 145	73.61 189	67.66 192	60.61 206	63.07 193

MVS-HIRNet		59.74 201	58.74 217	60.92 192	57.74 212	45.81 216	56.02 216	58.69 173	55.69 212	65.17 157	70.86 196	71.66 194	56.75 165	61.11 214	53.74 213	71.17 194	52.28 211

test_method		22.69 216	26.99 218	17.67 216	2.13 223	4.31 224	27.50 222	4.53 219	37.94 219	24.52 219	36.20 219	51.40 222	15.26 217	29.86 219	17.09 219	32.07 220	12.16 219

test123		1.06 217	1.41 219	0.64 218	0.39 224	0.48 225	0.52 227	0.25 222	1.11 223	1.37 226	2.01 222	1.98 228	0.87 220	1.43 221	1.27 220	0.46 224	1.62 221

testmvs		0.93 218	1.37 220	0.41 219	0.36 225	0.36 226	0.62 226	0.39 221	1.48 222	0.18 227	2.41 221	1.31 229	0.41 221	1.25 222	1.08 221	0.48 223	1.68 220

uanet_test		0.00 219	0.00 221	0.00 220	0.00 227	0.00 227	0.00 228	0.00 224	0.00 224	0.00 228	0.00 223	0.00 230	0.00 223	0.00 223	0.00 222	0.00 225	0.00 222

sosnet-low-res		0.00 219	0.00 221	0.00 220	0.00 227	0.00 227	0.00 228	0.00 224	0.00 224	0.00 228	0.00 223	0.00 230	0.00 223	0.00 223	0.00 222	0.00 225	0.00 222

sosnet		0.00 219	0.00 221	0.00 220	0.00 227	0.00 227	0.00 228	0.00 224	0.00 224	0.00 228	0.00 223	0.00 230	0.00 223	0.00 223	0.00 222	0.00 225	0.00 222

RE-MVS-def										87.10 28

9.14												89.43 130

SR-MVS					91.82 13			80.80 7				95.53 49

our_test_3					73.27 179	70.91 170	83.26 127

MTAPA										89.37 9		94.85 66

MTMP										90.54 5		95.16 59

Patchmatch-RL test							4.13 225

tmp_tt				13.54 217	16.73 222	6.42 223	8.49 224	2.36 220	28.69 221	27.44 218	18.40 220	13.51 227	3.70 219	33.23 218	36.26 218	22.54 222

XVS					91.28 25	91.23 8	96.89 2			87.14 25		94.53 71				95.84 15

X-MVStestdata					91.28 25	91.23 8	96.89 2			87.14 25		94.53 71				95.84 15

mPP-MVS					93.05 3							95.77 43

NP-MVS									78.65 130

Patchmtry						56.88 203	64.47 207	67.74 99		72.30 120

DeepMVS_CX						17.78 222	20.40 223	6.69 218	31.41 220	9.80 223	38.61 218	34.88 226	33.78 211	28.41 220		23.59 221	45.77 215

High-resolution multi-view benchmark