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		86.29 1	90.77 1	81.06 1	75.10 52	83.76 2	93.79 1	61.08 18	89.57 1	86.19 1	90.06 10	93.01 24	76.72 2	94.71 1	92.72 1	93.47 1	91.56 2

COLMAP_ROB	75.87 2	84.34 2	89.80 2	77.97 11	75.52 50	82.76 4	90.39 19	54.21 54	89.37 2	83.18 2	89.90 12	95.58 11	72.34 10	92.31 4	90.04 5	92.17 5	88.61 18
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
CP-MVS		84.06 3	86.79 10	80.86 2	81.81 8	79.66 29	92.67 6	64.48 1	83.13 30	82.32 3	80.89 73	92.97 25	72.51 9	91.74 6	90.02 6	91.40 17	89.14 8

ACMMPR		83.94 4	87.20 4	80.14 4	81.04 13	81.92 8	92.57 8	63.14 5	84.35 19	79.45 11	83.37 47	92.04 35	72.82 8	90.66 12	88.96 11	91.80 6	89.13 9

MP-MVS		83.50 5	86.11 19	80.45 3	82.58 4	80.60 24	92.68 5	63.48 3	81.43 44	80.21 9	81.95 63	90.76 53	72.86 6	90.14 19	89.30 10	90.92 19	88.59 19
Rongxuan Tan, Qing Wang, et al.: MP-MVS: Multi-Scale Windows PatchMatch and Planar Prior Multi-View Stereo.
ACMMP		83.17 6	86.75 11	79.01 7	80.11 25	82.01 7	92.29 10	60.35 25	82.20 38	78.32 15	80.59 74	93.14 22	70.67 15	91.30 8	89.36 9	92.30 4	88.62 17
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
PGM-MVS		83.03 7	85.67 26	79.95 5	80.69 17	81.09 15	92.40 9	63.06 6	79.38 59	80.21 9	80.31 76	91.44 40	71.75 12	90.46 15	88.53 14	91.57 9	88.50 20

LGP-MVS_train		82.91 8	86.50 13	78.72 8	78.72 34	81.03 16	89.78 23	61.16 17	80.15 54	80.44 6	84.83 37	94.19 14	70.52 17	90.70 11	87.19 22	91.71 8	87.37 29

ACMM	71.24 7	82.85 9	86.59 12	78.50 9	80.10 26	78.59 32	91.77 11	60.76 22	84.43 17	76.49 24	81.58 68	93.50 17	70.45 18	91.38 7	89.42 8	91.42 16	87.22 31
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
HFP-MVS		82.37 10	86.28 15	77.81 14	79.94 27	80.96 18	91.13 14	63.30 4	84.04 21	71.81 39	82.39 58	89.59 69	69.16 23	89.08 25	88.83 13	91.49 13	89.10 10

DeepC-MVS	73.80 3	82.34 11	86.87 8	77.06 18	78.62 35	84.34 1	90.30 21	63.54 2	83.10 31	71.30 44	86.91 24	90.54 59	67.12 29	87.81 34	87.05 23	91.46 15	88.37 21
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
CPTT-MVS		82.32 12	85.00 34	79.19 6	80.73 16	80.86 21	91.68 12	62.59 10	82.55 35	75.53 28	73.88 118	92.28 31	73.74 5	90.07 20	87.65 18	90.87 20	87.74 25

ACMP	70.35 9	82.17 13	86.45 14	77.18 17	79.33 28	81.00 17	89.27 27	58.63 31	81.35 46	75.46 29	82.97 53	95.08 12	68.90 24	90.49 14	87.43 21	91.48 14	86.84 33
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
SteuartSystems-ACMMP		82.16 14	85.55 28	78.21 10	80.48 19	79.28 30	92.65 7	61.03 19	80.55 52	77.00 22	81.80 66	90.71 54	68.73 25	90.25 17	87.94 17	89.36 27	88.30 22
Skip Steuart: Steuart Systems R&D Blog.
SMA-MVS		82.15 15	85.93 21	77.74 15	80.13 24	80.25 26	91.01 15	60.61 23	85.54 11	78.61 14	83.21 50	86.96 99	65.95 34	88.10 31	87.59 19	90.11 21	89.83 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
SD-MVS		82.13 16	86.80 9	76.67 19	80.36 22	80.66 22	89.48 25	56.93 34	82.50 36	67.55 67	87.05 22	91.40 42	72.84 7	88.66 27	88.32 15	92.85 2	89.04 11
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
LTVRE_ROB	75.99 1	82.04 17	87.16 5	76.07 22	63.57 124	70.27 74	86.48 41	62.99 7	89.00 5	80.32 7	86.25 27	91.04 47	74.66 4	92.58 3	90.29 4	88.42 34	90.72 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++		82.03 18	88.03 3	75.05 27	82.08 6	78.96 31	88.98 31	56.44 39	89.29 3	72.39 37	93.25 1	93.86 16	63.42 49	85.46 45	81.36 56	86.96 46	94.00 1

PMVS	70.37 8	81.82 19	87.08 6	75.68 24	77.06 42	77.23 40	87.77 38	56.25 42	83.33 29	67.18 72	89.48 15	87.94 83	77.70 1	93.02 2	92.57 2	88.13 36	86.00 38
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
ACMMP_NAP		81.79 20	85.72 24	77.21 16	79.15 32	79.68 28	91.62 13	59.66 27	83.55 26	77.74 18	83.72 45	87.34 92	65.36 35	88.61 28	87.56 20	89.73 26	89.58 6

X-MVS		81.61 21	84.73 36	77.97 11	80.31 23	81.29 12	93.53 2	62.50 11	81.41 45	77.45 19	72.04 130	90.19 62	62.50 56	90.57 13	88.87 12	91.54 10	88.73 15

OPM-MVS		81.44 22	85.68 25	76.49 20	79.27 29	78.21 35	89.84 22	58.67 30	85.25 12	76.26 25	85.28 34	92.88 26	66.03 33	87.20 37	85.40 27	88.86 31	85.58 42
Ray L. Khuboni, Hongjun Xu: Octagram Propagation Matching for Multi-Scale View Stereopsis (OPM-MVS).
TSAR-MVS + MP.		81.23 23	86.13 17	75.52 25	80.74 15	83.22 3	90.55 16	55.12 49	80.87 49	67.62 66	88.01 17	92.38 30	70.61 16	86.64 39	83.10 42	88.51 32	88.67 16
Zhenlong Yuan, Jiakai Cao, Zhaoqi Wang, Zhaoxin Li: TSAR-MVS: Textureless-aware Segmentation and Correlative Refinement Guided Multi-View Stereo. Pattern Recognition
TSAR-MVS + ACMM		81.20 24	86.92 7	74.52 29	77.60 38	82.29 5	84.41 48	62.95 8	82.99 32	64.03 81	87.71 18	89.17 72	71.98 11	88.19 30	88.10 16	86.18 55	89.95 4

APDe-MVS		81.08 25	86.12 18	75.20 26	79.25 30	80.91 19	90.38 20	57.05 33	85.83 10	66.07 77	87.34 21	91.27 43	69.45 19	85.99 43	82.55 44	88.98 30	88.95 13
Zhaojie Zeng, Yuesong Wang, Tao Guan: Matching Ambiguity-Resilient Multi-View Stereo via Adaptive Patch Deformation. Pattern Recognition
DPE-MVS		81.01 26	85.18 30	76.15 21	78.58 36	80.64 23	89.77 24	57.92 32	81.66 43	73.45 32	86.84 25	89.80 67	69.33 21	85.40 46	82.91 43	87.87 38	89.01 12
Kehua Chen, Zhenlong Yuan, Tianlu Mao, Zhaoqi Wang: Dual-Level Precision Edges Guided Multi-View Stereo with Accurate Planarization. AAAI2025
APD-MVS		80.60 27	84.63 37	75.91 23	81.22 11	81.48 10	90.49 17	58.81 29	77.54 65	67.49 68	85.90 29	89.82 66	69.43 20	86.08 42	83.80 37	88.01 37	87.77 24
Yuesong Wang, Zhaojie Zeng and etc.: Adaptive Patch Deformation for Textureless-Resilient Multi-View Stereo. CVPR2023
HPM-MVS++		80.44 28	82.57 49	77.96 13	81.99 7	72.76 61	90.48 18	61.31 14	80.85 50	77.90 17	81.93 64	87.01 96	68.20 27	84.15 56	85.27 29	87.85 39	86.00 38

DVP-MVS		80.31 29	85.60 27	74.15 33	76.23 46	78.39 33	86.62 40	55.79 45	86.47 9	71.32 43	90.96 7	89.02 74	69.28 22	84.62 55	81.64 53	85.66 60	88.09 23
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
ACMH+	67.97 10	80.15 30	86.16 16	73.14 40	73.82 59	76.41 43	83.59 52	54.82 52	87.35 6	70.86 48	86.98 23	96.27 4	66.50 30	89.17 24	83.39 39	89.26 28	83.56 48

OMC-MVS		79.95 31	85.28 29	73.74 36	72.95 62	80.10 27	87.87 37	48.13 84	84.62 16	79.42 12	80.27 77	92.49 28	64.14 43	87.25 36	85.11 30	89.92 24	87.10 32

SED-MVS		79.70 32	85.16 31	73.34 38	75.83 49	78.11 36	88.77 33	56.45 38	84.85 14	69.45 59	90.70 9	88.38 78	63.16 51	85.12 51	81.28 57	86.40 52	87.63 26

MSP-MVS		79.65 33	84.28 40	74.25 31	78.92 33	81.86 9	89.07 28	60.49 24	83.85 24	70.05 54	85.12 35	90.92 51	62.99 53	81.15 76	81.64 53	83.99 68	85.42 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
DeepPCF-MVS	71.57 5	79.49 34	84.05 41	74.17 32	74.14 56	80.88 20	89.33 26	56.24 43	82.41 37	71.58 41	82.27 59	86.47 102	66.47 31	84.80 53	84.16 35	87.26 44	87.34 30

LS3D		79.33 35	84.03 42	73.84 34	75.37 51	78.09 37	83.30 53	52.94 61	84.42 18	76.01 26	84.16 41	87.44 91	65.34 36	86.30 40	82.08 51	90.09 22	85.70 40

3Dnovator+	72.94 4	78.78 36	83.05 46	73.80 35	70.70 75	81.34 11	88.33 34	56.01 44	81.33 47	72.87 36	78.06 92	81.15 129	63.83 46	87.39 35	85.82 25	91.06 18	86.28 37

UA-Net		78.65 37	83.96 43	72.46 42	84.87 1	76.15 44	89.06 29	55.70 46	77.25 66	53.14 116	79.73 81	82.09 127	59.69 72	92.21 5	90.93 3	92.32 3	89.36 7

DeepC-MVS_fast	71.40 6	78.48 38	82.92 47	73.31 39	76.44 45	82.23 6	87.59 39	56.56 37	77.79 63	68.91 63	77.00 96	87.32 93	61.90 58	85.40 46	84.37 32	88.46 33	86.33 36
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
SF-MVS		78.36 39	83.47 45	72.41 43	76.04 48	75.72 47	83.86 50	51.81 65	84.00 23	70.65 51	81.27 70	92.22 32	64.64 41	83.28 65	80.28 59	87.44 43	87.49 28

WR-MVS		78.32 40	86.09 20	69.25 60	76.22 47	72.33 68	85.71 44	59.02 28	86.66 7	51.41 122	92.91 2	96.76 1	53.09 106	90.21 18	85.30 28	90.05 23	78.46 75

ACMH	66.19 11	78.12 41	84.55 38	70.63 51	69.62 81	72.40 67	80.77 68	46.43 96	89.24 4	77.99 16	87.42 20	95.83 9	62.95 54	86.27 41	78.24 70	86.00 58	82.46 50
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
train_agg		77.83 42	80.47 59	74.77 28	80.92 14	69.60 75	88.87 32	56.32 41	74.03 82	71.03 46	83.67 46	87.68 86	64.75 40	83.70 58	81.85 52	86.71 48	82.73 49

NCCC		77.82 43	80.72 58	74.43 30	79.24 31	75.72 47	88.06 35	56.36 40	79.61 56	73.22 34	67.75 143	87.05 95	63.09 52	85.62 44	84.00 36	86.62 49	85.30 45

CNVR-MVS		77.79 44	81.57 53	73.38 37	78.37 37	75.91 45	87.97 36	55.11 50	79.41 58	70.98 47	74.70 114	86.43 103	61.77 59	85.10 52	83.73 38	86.10 57	85.68 41

WR-MVS_H		77.56 45	85.88 22	67.86 64	80.54 18	74.32 55	83.23 54	61.78 12	83.47 27	47.46 142	91.81 6	95.84 8	50.50 118	90.44 16	84.37 32	83.63 73	80.89 61

RPSCF		77.56 45	84.51 39	69.46 59	65.17 112	74.36 54	79.74 75	47.45 87	84.01 22	72.89 35	77.89 93	90.67 55	65.14 38	88.25 29	89.74 7	86.38 53	86.64 35

PS-CasMVS		77.46 47	85.80 23	67.73 66	81.24 10	72.88 60	80.63 69	61.28 15	84.14 20	50.53 128	92.13 4	96.76 1	50.12 121	91.02 9	84.46 31	82.60 85	79.19 68

DTE-MVSNet		77.28 48	84.87 35	68.42 62	82.94 3	72.70 63	81.60 63	61.78 12	85.03 13	51.40 123	92.11 5	96.00 6	49.42 124	89.73 22	82.52 46	83.39 78	75.98 85

SixPastTwentyTwo		77.24 49	83.65 44	69.78 55	65.14 113	64.85 97	77.44 85	47.74 86	82.76 34	68.52 64	87.65 19	93.31 19	71.68 13	89.49 23	82.41 47	88.14 35	85.05 46

CDPH-MVS		77.22 50	81.05 57	72.75 41	77.29 40	77.46 39	86.36 42	54.02 56	73.00 88	69.75 57	77.78 94	88.90 76	61.31 63	84.09 57	82.54 45	87.79 40	83.57 47

PEN-MVS		77.06 51	85.05 32	67.74 65	82.29 5	72.59 64	80.86 67	61.03 19	84.66 15	50.08 131	92.19 3	96.59 3	49.12 126	89.83 21	82.35 48	83.06 79	77.14 80

CP-MVSNet		77.01 52	85.04 33	67.65 67	81.16 12	72.72 62	80.54 70	61.18 16	82.09 39	50.41 129	90.81 8	95.89 7	50.03 122	90.86 10	84.30 34	82.56 87	78.65 74

CSCG		76.95 53	82.08 51	70.97 47	73.32 61	78.35 34	81.08 66	47.19 88	83.47 27	69.82 56	80.44 75	87.19 94	64.59 42	81.01 79	77.26 77	89.83 25	86.84 33

CNLPA		76.67 54	81.72 52	70.77 50	70.75 73	76.68 42	86.14 43	46.11 98	81.82 41	74.68 30	76.37 98	86.23 105	62.92 55	85.28 49	83.29 40	84.02 67	82.40 51

MSLP-MVS++		76.66 55	82.32 50	70.06 53	70.51 76	80.27 25	79.77 74	55.58 47	77.79 63	63.09 82	67.25 147	89.50 70	71.01 14	88.10 31	85.74 26	80.39 99	87.56 27

TSAR-MVS + COLMAP		75.85 56	81.06 55	69.77 56	71.15 69	76.90 41	82.93 56	52.43 63	79.25 60	70.13 52	82.78 54	87.00 97	60.02 68	80.30 83	79.61 63	81.95 91	81.61 57

HQP-MVS		75.81 57	78.91 65	72.18 44	77.41 39	75.38 50	84.75 45	53.35 58	76.12 70	73.32 33	69.48 135	88.07 81	57.76 80	79.42 89	78.44 67	86.48 50	85.50 43

PLC	64.88 15	75.76 58	80.22 60	70.57 52	70.46 77	77.75 38	82.01 61	48.84 79	80.74 51	70.85 49	71.32 132	84.82 115	63.69 47	84.73 54	82.35 48	87.54 41	79.80 65
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
TAPA-MVS	66.11 12	75.37 59	79.24 63	70.86 48	67.63 88	74.09 56	83.17 55	44.75 113	81.82 41	80.83 5	65.61 158	88.04 82	61.58 60	83.21 66	80.12 60	87.17 45	81.82 54
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
PHI-MVS		75.17 60	78.37 66	71.43 45	71.13 70	72.46 66	82.28 60	50.55 71	73.39 85	79.05 13	73.65 120	87.50 90	61.98 57	81.10 77	78.48 66	83.60 74	81.99 52

anonymousdsp		74.76 61	82.59 48	65.63 84	45.61 207	61.13 124	89.06 29	32.58 193	74.11 81	59.55 91	84.06 42	94.12 15	75.24 3	88.94 26	86.95 24	91.74 7	88.81 14

AdaColmap		74.73 62	77.57 71	71.40 46	76.90 43	75.76 46	84.54 47	53.08 60	76.20 69	66.64 75	66.06 156	78.16 149	61.32 62	85.37 48	82.20 50	85.95 59	79.27 67

v7n		74.47 63	81.06 55	66.77 73	66.98 94	67.10 78	76.76 89	45.88 100	81.98 40	67.43 69	88.38 16	95.67 10	61.38 61	80.76 81	73.49 98	82.21 89	80.06 63

PCF-MVS	65.25 14	73.99 64	76.74 76	70.79 49	71.61 68	75.33 51	83.76 51	50.40 72	74.88 73	74.50 31	67.60 144	85.36 112	58.30 78	78.61 95	74.25 93	86.15 56	81.13 60
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
MCST-MVS		73.84 65	77.44 72	69.63 58	73.75 60	74.73 53	81.38 65	48.58 80	74.77 74	69.16 61	71.97 131	86.20 106	59.50 73	78.51 96	74.06 95	85.42 62	81.85 53

MVS_0304		73.74 66	77.16 74	69.74 57	74.24 55	73.47 57	84.70 46	49.62 74	62.26 144	67.27 70	75.87 102	87.57 88	57.49 83	81.20 75	79.50 64	85.10 63	80.27 62

TSAR-MVS + GP.		73.42 67	76.31 77	70.05 54	77.15 41	71.13 71	81.59 64	54.11 55	69.84 115	58.65 94	66.20 155	78.77 145	65.29 37	83.65 59	83.14 41	83.54 75	81.47 58

Gipuma		73.40 68	79.27 62	66.55 78	63.64 123	59.35 133	70.28 124	45.92 99	83.79 25	71.78 40	84.04 43	93.07 23	68.69 26	87.90 33	76.76 79	78.98 112	69.96 120
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
MVS_111021_HR		72.37 69	76.12 80	68.00 63	68.55 85	64.30 105	82.93 56	48.98 78	74.25 79	65.39 78	73.59 121	84.11 120	59.48 74	82.61 69	78.38 68	82.66 84	75.59 87

TinyColmap		71.85 70	76.11 81	66.87 72	66.07 103	65.34 92	74.35 101	49.30 77	79.93 55	75.93 27	75.66 104	87.74 85	54.72 98	80.66 82	70.42 120	80.85 97	73.02 103

UniMVSNet_ETH3D		71.84 71	81.36 54	60.74 108	76.46 44	66.01 86	66.49 144	60.24 26	86.58 8	41.87 165	90.04 11	96.02 5	43.72 153	85.14 50	77.30 76	75.64 133	68.40 134

TranMVSNet+NR-MVSNet		71.66 72	79.23 64	62.83 100	72.54 65	65.64 88	74.77 99	55.27 48	75.91 71	45.50 154	89.55 13	94.25 13	45.96 144	82.74 68	77.03 78	82.96 81	69.48 128

MVS_111021_LR		71.60 73	75.21 84	67.38 68	67.42 89	62.44 114	81.73 62	46.24 97	70.89 100	66.80 74	73.19 124	84.98 113	60.09 67	81.94 72	77.77 74	82.00 90	75.29 88

EG-PatchMatch MVS		71.50 74	76.82 75	65.30 85	70.74 74	66.50 82	74.23 103	43.25 121	72.02 92	59.11 92	79.85 80	86.88 100	63.95 45	80.29 84	75.25 89	80.51 98	76.98 81

DPM-MVS		71.35 75	73.50 102	68.84 61	74.93 53	73.35 58	84.07 49	50.56 70	71.91 93	67.06 73	61.21 178	77.02 154	52.64 109	74.15 116	75.14 90	83.79 71	81.74 55

UniMVSNet (Re)		71.29 76	78.14 67	63.30 95	70.29 78	66.57 81	75.98 91	54.74 53	70.20 108	46.20 152	85.08 36	93.21 20	48.19 132	82.50 70	78.33 69	84.40 65	71.08 114

CLD-MVS		71.24 77	78.12 68	63.20 97	74.03 57	71.60 69	82.82 58	32.91 190	74.23 80	69.32 60	79.65 82	91.54 38	47.02 140	81.22 74	79.01 65	73.09 150	69.63 124
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
CANet		71.07 78	75.09 86	66.39 79	72.57 64	71.53 70	82.38 59	47.10 89	59.81 150	59.81 90	74.97 109	84.37 119	54.25 101	79.89 87	77.64 75	82.25 88	77.40 78

v1192		71.06 79	74.87 88	66.61 75	66.38 98	65.80 87	78.27 79	45.28 105	70.19 109	70.79 50	83.37 47	91.79 36	58.76 77	70.86 151	69.02 127	80.16 102	73.08 101

DU-MVS		71.03 80	77.92 69	62.98 99	70.81 71	65.48 90	73.93 106	56.76 35	69.95 113	46.77 148	85.70 32	93.49 18	46.91 141	83.47 60	77.82 73	82.72 83	69.54 125

v1240		70.94 81	74.52 92	66.76 74	66.54 97	64.40 101	77.76 82	45.29 104	70.05 111	71.45 42	83.36 49	90.96 49	60.37 65	70.50 154	68.68 128	79.14 110	73.68 97

v1921920		70.82 82	74.46 94	66.58 77	66.33 99	64.35 104	77.72 83	45.07 107	70.39 105	71.18 45	83.15 51	90.62 57	59.97 69	70.90 149	68.43 135	79.19 109	73.39 98

UniMVSNet_NR-MVSNet		70.82 82	77.44 72	63.11 98	71.75 67	66.02 85	73.93 106	55.00 51	70.90 99	46.77 148	86.68 26	91.54 38	46.91 141	81.07 78	76.32 83	84.28 66	69.54 125

PVSNet_Blended_VisFu		70.70 84	73.62 101	67.28 70	63.53 125	72.96 59	77.97 80	52.10 64	63.65 138	62.66 83	71.14 133	73.46 165	63.55 48	79.35 93	75.34 88	83.90 69	79.43 66

v144192		70.68 85	74.40 96	66.34 80	65.94 105	64.38 102	77.63 84	45.18 106	69.97 112	70.11 53	82.70 56	90.77 52	59.84 71	71.43 147	68.46 131	79.31 108	73.08 101

EC-MVSNet		70.50 86	73.32 105	67.20 71	72.07 66	66.21 84	70.86 121	50.10 73	57.66 163	60.49 89	74.97 109	79.42 140	63.32 50	79.65 88	75.46 87	86.35 54	79.87 64

FPMVS		70.46 87	74.89 87	65.28 86	69.09 83	61.42 120	77.07 87	46.92 92	76.73 68	53.53 112	67.33 145	75.07 160	67.23 28	83.41 62	81.54 55	77.86 119	78.73 72

v1144		70.45 88	74.50 93	65.73 83	65.74 107	64.88 96	77.33 86	44.16 115	70.59 104	69.63 58	83.15 51	91.42 41	57.79 79	71.29 148	68.53 130	79.72 106	71.63 113

v10		70.25 89	74.59 91	65.19 87	65.32 110	66.46 83	76.60 90	44.84 110	67.38 124	67.21 71	82.75 55	90.56 58	57.70 81	71.69 144	68.63 129	79.44 107	74.67 91

Effi-MVS+-dtu		70.10 90	73.76 100	65.82 82	70.23 79	74.92 52	79.47 76	44.49 114	56.98 168	54.34 106	64.26 162	84.78 116	59.97 69	80.96 80	80.38 58	86.44 51	74.05 95

CS-MVS-test		70.00 91	72.92 108	66.59 76	66.71 96	64.06 107	80.28 72	44.82 111	58.41 155	58.08 97	73.57 122	80.94 131	63.98 44	83.29 64	75.93 85	85.65 61	74.23 92

MAR-MVS		70.00 91	72.28 116	67.34 69	69.89 80	72.57 65	80.09 73	49.49 76	60.28 149	69.03 62	59.29 187	80.79 133	54.68 99	78.39 98	76.00 84	80.87 96	78.67 73
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
Vis-MVSNet		69.95 93	77.69 70	60.91 106	60.67 140	66.71 79	77.94 81	48.58 80	69.10 118	45.78 153	80.21 78	83.58 123	53.41 105	82.92 67	80.11 61	79.08 111	81.21 59
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
CS-MVS		69.56 94	72.49 113	66.14 81	66.85 95	64.10 106	72.55 111	42.91 123	58.90 152	60.78 87	72.93 127	83.21 125	66.32 32	78.77 94	72.84 105	87.45 42	76.72 82

EPP-MVSNet		69.51 95	76.17 78	61.74 104	68.38 87	66.60 80	71.77 113	46.98 90	73.60 84	41.79 166	82.06 62	69.65 176	52.51 110	83.41 62	79.94 62	89.02 29	77.94 76

3Dnovator	65.69 13	69.43 96	75.74 83	62.06 103	60.78 139	70.50 73	75.85 93	39.57 158	74.44 76	57.41 99	75.91 100	77.73 151	55.34 94	76.86 102	75.61 86	83.44 76	79.14 69

Effi-MVS+		69.04 97	73.01 107	64.40 90	67.20 92	64.83 98	74.87 98	43.97 117	63.33 140	60.90 86	73.06 125	85.79 109	55.61 91	73.58 122	76.41 82	83.84 70	74.09 94

v2v482		69.01 98	73.39 104	63.89 92	63.86 117	62.99 111	75.26 96	42.05 132	70.22 107	68.46 65	82.64 57	91.61 37	55.38 92	70.89 150	66.93 151	78.30 115	68.48 133

MSDG		68.98 99	73.31 106	63.92 91	67.08 93	68.27 76	75.41 95	40.77 148	67.61 122	64.89 79	75.75 103	78.96 142	53.70 103	76.72 104	73.95 96	81.71 93	71.93 111

v8		68.77 100	73.50 102	63.26 96	63.74 121	64.47 100	74.22 104	42.07 131	67.30 125	64.89 79	82.08 61	90.23 61	56.50 89	71.85 143	66.57 152	78.14 116	72.02 109

NR-MVSNet		68.66 101	76.15 79	59.93 113	65.49 108	65.48 90	74.42 100	56.76 35	69.95 113	45.38 155	85.70 32	91.13 44	34.68 185	74.52 115	76.75 80	82.83 82	69.49 127

USDC		68.53 102	71.82 122	64.68 88	63.53 125	61.87 119	70.12 126	46.98 90	77.89 62	76.58 23	68.55 140	86.88 100	50.50 118	73.73 119	65.62 154	80.39 99	68.21 136

IS_MVSNet		68.20 103	74.41 95	60.96 105	68.55 85	64.36 103	71.47 116	48.33 82	70.11 110	43.30 161	80.90 72	74.54 162	47.19 139	81.25 73	77.97 72	86.94 47	71.76 112

Baseline_NR-MVSNet		68.15 104	75.12 85	60.02 112	70.81 71	55.67 156	75.88 92	53.40 57	71.25 96	43.96 159	85.88 30	92.68 27	45.76 145	83.47 60	68.34 136	70.34 167	68.58 132

GeoE		68.11 105	72.10 120	63.47 94	67.32 90	62.42 115	78.32 78	43.22 122	64.06 137	55.72 104	73.97 117	84.58 117	55.35 93	76.09 108	70.41 121	80.89 95	73.14 100

casdiffmvs_mvg		68.01 106	74.22 98	60.78 107	63.75 120	62.08 118	70.21 125	42.58 125	72.11 91	58.53 95	84.80 39	88.98 75	49.37 125	73.25 126	67.54 147	80.24 101	70.75 116
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
Fast-Effi-MVS+		67.71 107	72.54 112	62.07 102	63.83 118	63.68 108	75.74 94	39.94 154	60.89 148	54.29 107	73.00 126	86.19 107	56.85 86	78.46 97	73.23 101	81.74 92	72.36 107

thisisatest0515		66.95 108	72.29 115	60.72 109	56.37 164	56.05 154	71.08 117	38.81 163	67.59 123	53.26 115	78.21 90	79.79 139	60.11 66	75.69 110	73.02 103	84.69 64	75.66 86

EPNet		66.87 109	68.89 137	64.53 89	73.97 58	61.13 124	78.46 77	61.03 19	56.78 170	53.41 113	66.91 150	70.91 171	43.49 154	76.08 109	76.68 81	76.81 123	73.73 96
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
canonicalmvs		66.37 110	74.37 97	57.04 130	65.89 106	65.06 93	62.58 163	42.55 126	76.82 67	46.87 147	67.33 145	86.38 104	45.49 147	76.77 103	71.85 110	78.87 113	76.35 83

QAPM		66.36 111	72.76 111	58.90 118	59.57 147	65.01 94	64.05 157	41.17 143	73.09 87	56.82 101	69.42 136	77.78 150	55.07 96	73.00 129	72.07 108	76.71 124	78.96 70

casdiffmvs		66.19 112	72.34 114	59.02 117	62.75 129	60.61 130	69.06 131	41.38 140	69.49 116	54.11 108	84.00 44	89.74 68	49.12 126	70.74 153	62.70 169	77.70 121	69.14 130
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
V42		65.79 113	72.11 119	58.42 122	51.89 186	58.69 135	73.80 108	34.50 182	65.40 132	57.10 100	79.54 84	89.09 73	57.51 82	71.98 141	67.83 144	75.70 131	72.26 108

IterMVS-LS		65.76 114	70.85 130	59.81 115	65.33 109	57.78 140	64.63 154	48.02 85	65.65 131	51.05 125	81.31 69	77.47 152	54.94 97	69.46 162	69.36 126	74.90 139	74.95 89
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
PM-MVS		65.66 115	71.25 128	59.14 116	58.92 153	54.88 165	73.66 110	38.55 166	66.12 128	49.91 133	69.87 134	86.97 98	60.61 64	76.30 106	74.75 91	73.19 148	69.83 121

UGNet		65.61 116	74.79 89	54.91 142	54.54 180	68.20 77	70.97 120	48.21 83	67.14 127	41.67 167	74.15 116	80.65 135	36.10 180	79.39 90	77.99 71	77.95 118	76.01 84
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
DELS-MVS		65.54 117	71.79 123	58.24 125	59.68 146	65.55 89	70.99 118	38.69 165	62.29 143	49.27 136	75.03 108	81.42 128	50.93 114	73.71 121	71.35 112	79.90 104	73.20 99
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
pmmvs-eth3d		65.36 118	70.09 133	59.85 114	63.05 128	53.61 168	74.29 102	46.45 95	68.14 121	51.45 121	78.83 88	85.78 110	49.87 123	70.44 155	70.45 119	74.00 143	63.38 156

FA-MVS(training)		65.22 119	69.19 136	60.58 110	60.91 137	57.60 144	71.57 115	38.82 162	65.84 130	61.18 85	75.95 99	74.52 163	51.18 113	73.32 124	67.41 148	79.80 105	69.70 123

v148		64.92 120	70.58 132	58.32 123	59.89 144	57.09 147	66.04 146	35.27 181	69.11 117	60.66 88	79.57 83	90.93 50	53.91 102	69.81 161	62.22 170	74.14 141	65.31 148

FC-MVSNet-train		64.87 121	74.76 90	53.33 146	65.24 111	58.05 137	69.69 128	41.92 135	70.99 98	32.62 191	85.75 31	88.23 79	32.10 196	77.61 101	74.41 92	78.43 114	68.25 135

pmmvs6		64.78 122	75.82 82	51.89 152	62.41 130	57.13 146	60.24 170	45.59 102	82.90 33	34.69 182	84.83 37	93.18 21	36.22 179	76.43 105	71.13 115	72.21 154	65.12 149

OpenMVS	60.79 16	64.42 123	69.72 134	58.23 126	61.63 134	62.17 116	64.11 156	37.54 175	67.17 126	55.71 105	65.89 157	74.89 161	52.67 108	72.20 139	68.29 138	77.73 120	77.39 79

test1111		64.34 124	71.57 124	55.90 137	67.25 91	60.24 132	66.66 142	51.63 68	73.36 86	34.69 182	75.63 105	80.67 134	39.43 165	78.17 99	71.69 111	75.71 130	61.23 162

DCV-MVSNet		64.34 124	72.84 110	54.42 144	63.79 119	62.09 117	62.50 164	42.72 124	74.32 78	41.34 168	66.96 148	88.57 77	39.18 166	75.20 112	70.35 122	77.01 122	72.37 106

ETV-MVS		64.30 126	64.76 152	63.77 93	68.59 84	62.49 113	77.02 88	45.31 103	49.27 196	50.88 126	56.23 199	59.91 199	57.12 85	80.19 86	74.23 94	83.68 72	71.03 115

ECVR-MVS		63.93 127	71.52 125	55.08 140	66.19 100	61.34 121	63.84 158	51.79 66	70.75 102	34.77 180	74.70 114	81.10 130	38.92 167	79.39 90	73.43 99	75.00 137	59.92 169

Anonymous20231211		63.69 128	72.86 109	53.00 149	63.72 122	60.25 131	60.33 169	40.96 144	72.49 89	38.91 171	81.77 67	88.17 80	37.60 174	73.30 125	68.01 141	76.47 128	66.06 145

TransMVSNet (Re)		63.49 129	73.86 99	51.39 158	64.26 116	56.07 153	61.17 167	42.23 129	78.81 61	34.80 179	85.94 28	90.63 56	34.35 189	72.73 134	67.98 142	71.50 157	64.84 150

DI_MVS_plusplus_trai		63.43 130	67.54 141	58.63 119	62.34 131	58.06 136	65.75 150	42.15 130	63.05 141	53.28 114	75.88 101	75.92 157	50.18 120	68.04 165	64.20 160	78.07 117	67.65 137

EIA-MVS		63.24 131	64.16 156	62.16 101	69.30 82	63.20 110	72.40 112	40.82 147	48.31 202	51.50 120	59.63 185	62.23 191	57.33 84	78.00 100	71.94 109	81.59 94	65.82 146

Fast-Effi-MVS+-dtu		63.22 132	65.55 146	60.49 111	61.24 136	64.70 99	74.15 105	53.24 59	51.46 181	49.67 134	58.03 194	78.42 146	48.05 134	72.03 140	71.14 114	76.60 127	63.09 157

IterMVS-SCA-FT		62.67 133	68.00 139	56.45 136	56.92 162	64.92 95	57.51 183	38.12 167	59.44 151	53.62 111	74.74 113	71.60 168	64.84 39	70.24 157	65.27 156	67.70 175	69.83 121

diffmvs		62.64 134	69.66 135	54.46 143	56.19 167	55.06 162	67.36 140	36.74 179	64.18 135	50.58 127	79.54 84	87.55 89	45.13 149	68.04 165	63.20 165	70.78 162	70.02 119
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
MVS_Test		62.58 135	67.46 142	56.89 132	59.52 148	55.90 155	64.94 152	38.83 161	57.08 167	56.55 102	76.53 97	84.49 118	47.45 136	66.95 168	62.01 171	74.04 142	69.27 129

MDA-MVSNet-bldmvs		62.46 136	72.13 118	51.19 160	34.32 217	56.10 152	68.65 133	38.85 160	69.05 119	49.50 135	78.17 91	85.43 111	51.32 111	86.67 38	67.40 149	64.46 180	62.08 159

pm-mvs1		61.97 137	72.01 121	50.25 166	60.64 141	55.23 160	58.67 178	42.44 127	74.40 77	33.63 188	81.03 71	89.86 65	34.87 184	72.93 132	67.95 143	71.28 158	62.65 158

FMVSNet1		61.92 138	71.36 126	50.90 162	57.67 161	59.29 134	59.48 176	44.14 116	70.24 106	34.72 181	75.45 107	84.94 114	36.75 176	72.33 136	68.45 132	72.66 151	68.83 131

PVSNet_BlendedMVS		61.75 139	65.07 150	57.87 128	56.27 165	60.99 127	65.81 148	43.75 118	51.27 185	54.08 109	62.12 173	78.84 143	50.67 115	71.49 145	63.91 162	76.64 125	66.86 139

PVSNet_Blended		61.75 139	65.07 150	57.87 128	56.27 165	60.99 127	65.81 148	43.75 118	51.27 185	54.08 109	62.12 173	78.84 143	50.67 115	71.49 145	63.91 162	76.64 125	66.86 139

tttt0517		61.44 141	63.85 158	58.62 120	55.20 176	55.61 157	68.80 132	38.02 170	55.70 172	50.01 132	66.93 149	48.90 210	56.69 87	73.84 118	71.10 116	82.99 80	74.89 90

tfpnnormal		61.41 142	71.33 127	49.83 167	61.73 133	54.90 164	58.52 179	41.24 141	75.20 72	32.00 196	82.13 60	87.87 84	35.63 183	72.75 133	66.30 153	69.87 168	60.14 166

IB-MVS	57.02 17	61.37 143	65.39 147	56.69 133	56.65 163	60.85 129	70.70 122	37.90 172	49.37 195	45.37 156	48.75 211	79.14 141	53.55 104	76.26 107	70.85 118	75.97 129	72.50 105
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
CANet_DTU		61.22 144	67.07 143	54.40 145	59.89 144	63.62 109	70.98 119	36.77 178	50.49 188	47.15 143	62.45 171	80.81 132	37.90 173	71.87 142	70.09 124	73.69 144	70.19 118

pmmvs4		61.12 145	64.61 153	57.04 130	60.88 138	52.15 174	70.59 123	44.82 111	61.35 147	46.91 146	72.08 129	73.27 166	46.79 143	65.06 172	67.76 145	72.28 152	60.58 164

thisisatest0530		61.02 146	63.44 163	58.19 127	54.75 178	55.09 161	68.03 137	38.02 170	55.45 173	49.06 137	66.58 153	48.69 211	56.69 87	73.07 127	71.10 116	82.60 85	74.14 93

Vis-MVSNet (Re-imp)		60.99 147	67.78 140	53.06 148	64.66 115	53.49 169	67.40 138	49.52 75	68.55 120	28.00 205	79.53 86	71.41 170	33.08 193	75.30 111	71.28 113	75.69 132	54.91 181

PatchMatch-RL		60.96 148	63.00 166	58.57 121	59.16 152	52.18 173	67.38 139	41.99 133	57.85 161	48.16 138	53.55 207	69.77 175	59.47 75	73.73 119	72.49 107	75.27 136	61.44 161

GA-MVS		60.73 149	64.24 155	56.64 134	59.38 151	57.45 145	65.07 151	36.65 180	57.39 165	58.17 96	73.43 123	69.10 179	47.38 137	64.47 177	63.63 164	73.19 148	64.22 153

CVMVSNet		60.45 150	63.72 160	56.63 135	54.82 177	53.75 166	68.41 134	41.95 134	55.07 174	48.03 139	58.08 193	68.67 180	55.09 95	69.14 163	68.34 136	71.51 156	72.97 104

ET-MVSNet_ETH3D		60.33 151	62.10 169	58.27 124	58.61 156	58.05 137	68.06 135	41.20 142	51.40 182	51.10 124	64.06 163	49.42 209	50.61 117	74.72 113	70.29 123	80.05 103	66.74 141

FC-MVSNet-test		60.28 152	70.83 131	47.96 178	54.69 179	47.12 188	68.06 135	41.68 139	71.42 94	23.73 213	84.70 40	77.41 153	28.92 201	82.33 71	73.08 102	70.68 163	59.77 170

test2506		59.86 153	64.01 157	55.02 141	66.19 100	61.34 121	63.84 158	51.79 66	70.75 102	34.39 186	62.65 170	39.92 225	38.92 167	79.39 90	73.43 99	75.00 137	60.56 165

EU-MVSNet		59.77 154	66.07 144	52.42 150	47.81 196	51.86 176	62.98 162	32.28 195	62.08 145	47.10 144	59.94 183	83.42 124	53.08 107	70.06 160	63.19 166	71.26 160	71.96 110

IterMVS		59.24 155	64.45 154	53.16 147	50.98 188	61.29 123	66.51 143	32.85 191	58.17 157	46.31 151	72.58 128	70.23 173	54.26 100	64.81 175	60.24 176	68.04 174	63.81 155
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
HyFIR lowres test		59.15 156	62.28 168	55.49 139	52.42 184	62.59 112	71.76 114	39.74 155	50.25 190	41.92 164	62.88 168	69.16 178	55.85 90	62.77 182	67.18 150	71.27 159	61.11 163

thres600view7		58.87 157	65.91 145	50.66 164	61.27 135	56.32 150	59.88 174	40.63 150	64.88 133	32.10 195	64.82 159	69.83 174	36.72 177	72.99 130	72.55 106	73.34 146	59.97 168

CMPMVS	45.32 18	58.10 158	65.24 149	49.76 168	47.88 195	46.86 191	48.16 209	32.82 192	58.06 158	61.35 84	59.64 184	80.00 136	47.27 138	70.15 158	64.10 161	61.08 184	77.85 77
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
MDTV_nov1_ep13_2view		58.09 159	63.54 162	51.74 154	50.13 190	46.56 192	66.95 141	33.41 188	63.52 139	58.77 93	74.84 111	84.10 121	43.12 155	65.95 171	54.69 191	58.04 190	55.13 180

CDS-MVSNet		57.90 160	63.57 161	51.28 159	62.30 132	53.17 170	64.70 153	51.61 69	57.41 164	32.75 190	63.73 164	70.53 172	27.12 203	72.49 135	73.02 103	69.22 171	54.68 182
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
FMVSNet2		57.80 161	65.39 147	48.94 174	55.88 168	57.61 141	57.26 184	42.37 128	58.21 156	33.19 189	68.36 141	75.55 159	34.58 186	66.91 169	64.55 158	70.38 164	66.56 142

thres400		57.25 162	63.73 159	49.70 169	60.19 143	54.95 163	58.16 180	39.60 157	62.42 142	31.98 198	62.33 172	69.20 177	35.96 181	70.07 159	68.03 140	72.28 152	59.12 172

WB-MVS		57.14 163	71.03 129	40.93 196	47.97 194	53.10 171	52.35 199	39.66 156	79.59 57	23.31 214	89.49 14	89.23 71	32.94 194	74.68 114	61.41 173	49.32 206	49.74 191

gm-plane-assit		56.76 164	57.64 183	55.73 138	66.01 104	55.45 159	74.96 97	30.54 200	73.71 83	56.04 103	81.81 65	30.91 228	43.83 151	58.77 192	54.71 190	63.02 182	48.13 198

MIMVSNet1		56.72 165	68.69 138	42.76 192	46.70 202	42.81 199	69.13 129	30.52 201	81.01 48	32.00 196	74.82 112	91.10 46	26.83 205	73.98 117	64.72 157	51.40 202	52.38 186

EPNet_dtu		56.63 166	60.77 175	51.80 153	55.47 173	44.63 193	69.83 127	38.74 164	50.27 189	47.64 140	58.01 195	72.27 167	33.71 191	68.60 164	67.72 146	65.39 178	63.86 154
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
GBi-Net		56.54 167	63.26 164	48.70 175	55.88 168	57.61 141	57.26 184	41.75 136	49.06 197	32.37 192	61.81 175	67.02 182	34.58 186	72.33 136	68.45 132	70.38 164	66.56 142

test1		56.54 167	63.26 164	48.70 175	55.88 168	57.61 141	57.26 184	41.75 136	49.06 197	32.37 192	61.81 175	67.02 182	34.58 186	72.33 136	68.45 132	70.38 164	66.56 142

gg-mvs-nofinetune		56.45 169	61.04 173	51.10 161	63.42 127	49.40 184	53.71 195	52.52 62	74.77 74	46.93 145	77.31 95	53.88 203	26.42 207	62.51 183	57.81 183	63.60 181	51.57 189

thres200		56.35 170	62.36 167	49.34 171	58.87 154	56.32 150	55.91 188	40.63 150	58.51 154	31.34 199	58.81 191	67.31 181	35.96 181	72.99 130	65.51 155	73.34 146	57.07 175

MS-PatchMatch		56.31 171	60.22 178	51.73 155	60.53 142	55.53 158	63.41 160	37.18 176	51.34 184	37.44 173	60.53 181	62.19 192	45.52 146	64.25 178	63.17 167	66.33 176	64.56 151

tfpn200view9		56.07 172	61.85 170	49.34 171	58.57 157	56.48 149	58.01 182	40.72 149	53.23 175	31.01 200	56.41 197	66.40 187	34.18 190	73.02 128	68.06 139	73.53 145	59.35 171

FMVSNet3		54.77 173	60.73 176	47.81 179	54.29 181	56.88 148	55.89 189	41.75 136	49.06 197	32.37 192	61.81 175	67.02 182	33.67 192	62.88 181	61.96 172	68.88 172	65.53 147

thres100view900		53.88 174	59.19 179	47.68 180	58.57 157	52.74 172	54.45 192	38.07 169	53.23 175	31.01 200	56.41 197	66.40 187	32.80 195	65.03 174	64.43 159	71.18 161	56.10 178

CR-MVSNet		53.82 175	55.40 187	51.98 151	51.57 187	50.23 179	45.00 211	44.97 108	46.90 204	52.60 117	67.91 142	46.99 218	48.37 129	59.15 190	59.53 179	69.38 170	57.07 175

baseline2		53.55 176	55.19 188	51.62 156	55.27 175	51.95 175	60.89 168	34.23 183	46.69 206	42.47 162	53.56 206	50.01 206	45.33 148	64.63 176	61.22 174	71.56 155	58.28 174

test20.03		53.49 177	60.95 174	44.78 188	64.73 114	47.25 187	61.58 166	43.30 120	65.86 129	22.85 215	66.87 152	79.85 137	22.99 209	62.38 184	56.95 185	53.25 199	47.46 200

baseline		53.46 178	61.55 171	44.01 189	45.83 205	48.77 185	57.26 184	28.75 204	49.99 191	38.85 172	68.78 138	75.65 158	38.30 170	60.80 185	59.78 178	55.10 197	67.07 138

MVSTER		53.08 179	56.39 185	49.21 173	47.19 198	51.08 177	60.14 172	31.74 197	40.63 215	38.97 170	55.78 200	46.74 219	42.47 158	67.29 167	62.99 168	74.73 140	70.23 117

CHOSEN 1792x2688		52.99 180	56.91 184	48.42 177	47.32 197	50.10 181	64.18 155	33.85 185	45.46 209	36.95 175	55.20 203	66.49 186	51.20 112	59.28 188	59.81 177	57.01 192	61.99 160

baseline1		52.90 181	58.38 180	46.51 186	58.87 154	50.01 182	54.17 193	40.45 153	56.81 169	29.25 203	62.72 169	58.99 200	30.25 199	65.05 173	60.57 175	66.07 177	54.54 183

CostFormer		52.59 182	55.14 189	49.61 170	52.72 182	50.40 178	66.28 145	33.78 186	52.85 177	43.43 160	66.30 154	51.37 205	41.78 161	54.92 202	51.18 198	59.68 186	58.98 173

SCA		52.47 183	53.97 192	50.71 163	46.95 201	57.79 139	60.18 171	46.89 93	51.92 180	46.71 150	60.73 179	49.97 207	47.69 135	56.39 199	52.98 195	55.82 194	48.03 199

testgi		51.94 184	61.37 172	40.94 195	58.38 159	47.03 189	65.88 147	30.49 202	70.87 101	22.64 216	57.53 196	87.59 87	18.30 216	63.01 180	54.32 192	49.93 205	49.27 193

dmvs_re		51.01 185	54.88 190	46.49 187	58.06 160	44.35 196	60.08 173	37.67 173	42.11 213	28.68 204	45.12 216	66.70 185	31.90 198	66.62 170	59.18 181	62.59 183	60.11 167

tpm cat1		50.98 186	51.28 198	50.62 165	55.74 171	49.92 183	63.13 161	38.12 167	52.38 179	47.61 141	60.11 182	44.51 222	44.86 150	51.31 210	47.49 205	54.25 198	53.24 185

RPMNet		50.92 187	50.32 200	51.62 156	50.25 189	50.23 179	59.16 177	46.70 94	46.90 204	42.39 163	48.97 210	37.23 226	41.78 161	57.30 197	56.18 187	69.44 169	55.43 179

pmmvs5		50.64 188	58.01 181	42.05 193	47.01 200	43.67 197	49.27 207	29.43 203	50.77 187	33.83 187	68.69 139	76.16 156	27.82 202	57.53 196	57.07 184	64.95 179	52.18 187

PatchT		50.55 189	53.55 194	47.05 184	37.59 216	42.26 200	50.55 204	37.56 174	46.37 207	52.60 117	66.91 150	43.54 224	48.37 129	59.15 190	59.53 179	55.62 195	57.07 175

Anonymous20231206		50.28 190	57.94 182	41.35 194	55.45 174	43.65 198	58.06 181	34.12 184	62.02 146	24.25 212	59.33 186	79.80 138	24.49 208	59.55 186	54.28 193	51.74 201	46.94 202

dps		49.71 191	51.97 196	47.07 183	52.37 185	47.00 190	53.02 198	40.52 152	44.91 210	41.23 169	64.55 160	44.27 223	40.12 164	57.71 195	51.97 197	55.14 196	53.41 184

MDTV_nov1_ep13		49.60 192	51.57 197	47.31 181	46.28 203	44.61 194	59.82 175	30.96 198	48.80 201	50.20 130	59.26 188	52.38 204	38.56 169	56.20 200	49.70 201	58.04 190	50.01 190

PatchmatchNet		48.67 193	50.10 201	46.99 185	48.29 193	41.00 201	55.54 190	38.94 159	51.38 183	45.15 157	63.22 166	48.45 213	42.83 156	53.80 207	48.50 204	51.19 204	44.37 204
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
new-patchmatchnet		47.33 194	60.49 177	31.99 212	55.69 172	33.86 213	36.84 221	33.31 189	72.36 90	14.33 221	80.09 79	92.14 33	13.27 218	63.54 179	40.09 213	38.51 215	41.32 212

tpm		46.67 195	49.20 206	43.72 190	49.60 191	36.60 210	53.93 194	26.84 206	52.70 178	58.05 98	69.04 137	47.96 214	30.06 200	48.33 213	42.76 208	43.88 209	47.01 201

pmmvs3		46.64 196	54.13 191	37.90 203	31.23 221	40.68 202	49.83 206	15.34 217	46.31 208	36.34 177	53.15 208	74.40 164	36.36 178	58.43 193	56.64 186	58.32 189	49.29 192

TAMVS		46.64 196	53.62 193	38.49 201	49.56 192	36.87 207	53.16 197	25.76 208	56.33 171	22.55 217	60.72 180	61.80 194	27.12 203	59.50 187	58.33 182	52.79 200	41.82 211

test-LLR		46.01 198	45.06 215	47.11 182	59.39 149	36.72 208	51.28 201	40.95 145	36.41 219	34.45 184	46.14 213	47.02 216	38.00 171	51.78 208	48.53 202	58.60 187	48.84 195

MIMVSNet		45.83 199	53.46 195	36.94 204	45.38 209	39.50 204	52.20 200	30.68 199	57.09 166	24.53 211	55.22 202	71.54 169	21.74 212	55.81 201	51.08 199	47.11 207	43.96 205

pmnet_mix02		45.67 200	55.99 186	33.63 211	45.77 206	31.22 217	42.04 216	27.60 205	64.14 136	24.89 208	75.50 106	82.30 126	21.88 211	54.53 205	41.22 210	39.62 213	43.05 207

test0.0.03 1		45.40 201	49.55 204	40.57 198	59.39 149	44.36 195	53.37 196	40.95 145	47.14 203	19.23 218	45.49 215	60.24 197	19.24 214	54.82 203	51.98 196	51.21 203	42.82 208

PMMVS		45.37 202	49.29 205	40.79 197	27.75 222	35.07 212	50.88 203	19.88 214	39.27 217	35.78 178	50.11 209	61.29 195	42.04 159	54.13 206	55.95 188	68.43 173	49.19 194

MVS-HIRNet		44.56 203	45.52 213	43.44 191	40.98 212	31.03 218	39.52 220	36.96 177	42.80 212	44.37 158	53.80 205	60.04 198	41.85 160	47.97 215	41.08 211	56.99 193	41.95 210

test-mter		44.18 204	47.60 208	40.18 199	33.20 218	39.03 205	55.28 191	13.91 219	39.07 218	36.63 176	48.09 212	49.52 208	41.12 163	54.55 204	50.91 200	60.97 185	52.03 188

EMVS		43.85 205	49.91 202	36.77 206	45.46 208	32.70 214	44.09 213	25.33 209	57.88 160	26.62 206	58.99 190	61.14 196	42.77 157	70.26 156	38.52 217	36.38 217	29.87 218

E-PMN		43.83 206	49.81 203	36.84 205	46.09 204	31.86 216	42.77 215	25.85 207	57.76 162	25.53 207	55.50 201	62.47 190	43.77 152	70.78 152	39.51 214	37.04 216	30.79 217

tpmrst		43.31 207	46.14 211	40.02 200	47.05 199	36.48 211	48.01 210	32.17 196	49.50 194	37.26 174	63.66 165	47.04 215	31.98 197	42.00 219	40.55 212	43.64 210	43.75 206

TESTMET0.1,1		41.79 208	45.06 215	37.97 202	31.32 220	36.72 208	51.28 201	14.17 218	36.41 219	34.45 184	46.14 213	47.02 216	38.00 171	51.78 208	48.53 202	58.60 187	48.84 195

ADS-MVSNet		40.61 209	46.31 209	33.96 209	40.70 213	30.42 219	40.42 218	33.44 187	58.01 159	30.87 202	63.05 167	54.48 202	22.67 210	44.35 218	39.23 216	35.64 218	34.64 215

CHOSEN 280x420		40.24 210	44.14 218	35.69 207	32.36 219	23.58 222	50.30 205	21.21 213	40.94 214	18.84 219	32.75 220	48.65 212	48.13 133	59.16 189	55.31 189	43.28 211	48.62 197

EPMVS		40.11 211	44.96 217	34.44 208	41.55 211	32.65 215	41.74 217	32.39 194	49.89 193	24.83 209	64.44 161	46.38 220	26.57 206	44.75 217	39.47 215	39.59 214	37.16 214

FMVSNet5		39.83 212	45.08 214	33.71 210	39.24 214	39.56 203	48.77 208	23.55 211	39.45 216	24.55 210	33.73 219	44.57 221	20.97 213	58.27 194	54.23 194	45.16 208	45.77 203

N_pmnet		39.50 213	51.01 199	26.09 214	44.48 210	25.59 221	40.20 219	21.49 212	64.20 134	7.98 225	73.86 119	76.67 155	13.66 217	50.17 211	36.69 219	28.71 220	29.86 219

MVE	28.01 19	35.86 214	43.56 219	26.88 213	22.33 224	19.75 224	30.85 224	23.88 210	49.90 192	10.48 223	43.64 217	61.87 193	48.99 128	47.26 216	42.15 209	24.76 221	40.37 213
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
new_pmnet		35.76 215	45.64 212	24.22 215	38.59 215	25.83 220	31.87 223	19.24 215	49.06 197	9.01 224	54.34 204	64.73 189	12.46 219	49.21 212	44.91 206	34.17 219	31.41 216

PMMVS2		34.11 216	48.55 207	17.26 216	25.45 223	20.72 223	35.08 222	16.26 216	58.71 153	4.16 227	59.22 189	78.40 147	3.65 221	57.24 198	38.31 218	18.94 223	27.28 220

GG-mvs-BLEND		31.54 217	46.27 210	14.37 217	0.07 229	48.65 186	42.97 214	0.08 225	44.04 211	1.21 229	39.77 218	57.94 201	0.15 225	48.19 214	42.82 207	41.70 212	42.46 209

test_method		13.28 218	15.83 220	10.30 218	1.05 226	2.18 227	15.40 225	2.23 221	22.43 221	13.84 222	22.00 222	33.14 227	9.78 220	17.80 221	9.93 221	19.50 222	3.31 222

test123		0.53 219	0.60 222	0.46 220	0.22 227	0.25 228	0.33 230	0.13 224	0.66 224	1.37 228	1.10 224	0.00 232	0.43 223	0.68 223	0.61 222	0.26 226	0.88 223

testmvs		0.47 220	0.69 221	0.21 221	0.17 228	0.17 229	0.35 229	0.16 223	0.66 224	0.18 230	1.05 225	0.99 231	0.27 224	0.62 224	0.54 223	0.15 227	0.77 224

uanet_test		0.00 221	0.00 223	0.00 222	0.00 230	0.00 230	0.00 231	0.00 226	0.00 226	0.00 231	0.00 226	0.00 232	0.00 226	0.00 225	0.00 224	0.00 228	0.00 225

sosnet-low-res		0.00 221	0.00 223	0.00 222	0.00 230	0.00 230	0.00 231	0.00 226	0.00 226	0.00 231	0.00 226	0.00 232	0.00 226	0.00 225	0.00 224	0.00 228	0.00 225

sosnet		0.00 221	0.00 223	0.00 222	0.00 230	0.00 230	0.00 231	0.00 226	0.00 226	0.00 231	0.00 226	0.00 232	0.00 226	0.00 225	0.00 224	0.00 228	0.00 225

TPM-MVS					72.72 63	70.92 72	80.38 71			66.22 76	58.35 192	78.23 148	48.26 131			83.40 77	81.74 55
Ray Leroy Khuboni and Hongjun Xu: Textureless Resilient Propagation Matching in Multiple View Stereosis (TPM-MVS). SATNAC 2025
RE-MVS-def										70.04 55

9.14												83.64 122

SR-MVS					81.31 9			62.63 9				91.11 45

Anonymous202405211			72.22 117		66.19 100	61.09 126	62.23 165	45.87 101	71.25 96		79.33 87	86.16 108	37.36 175	73.54 123	69.84 125	75.45 134	64.32 152

our_test_3					52.72 182	53.66 167	69.11 130

ambc			79.96 61		74.57 54	75.48 49	73.75 109		80.32 53	72.34 38	78.46 89	92.41 29	59.05 76	80.24 85	73.95 96	75.41 135	78.85 71

MTAPA										80.26 8		90.53 60

MTMP										82.07 4		91.00 48

Patchmatch-RL test							2.05 228

tmp_tt				7.47 219	8.89 225	3.32 226	4.35 227	1.14 222	15.58 223	15.76 220	8.50 223	5.90 230	2.00 222	20.02 220	21.51 220	12.70 224

XVS					80.47 20	81.29 12	93.33 3			77.45 19		90.19 62				91.52 11

X-MVStestdata					80.47 20	81.29 12	93.33 3			77.45 19		90.19 62				91.52 11

mPP-MVS					82.97 2							92.12 34

NP-MVS									71.39 95

Patchmtry						37.73 206	45.00 211	44.97 108		52.60 117

DeepMVS_CX						8.52 225	9.75 226	3.19 220	16.70 222	5.02 226	23.06 221	19.33 229	18.69 215	13.75 222		11.34 225	25.07 221

High-resolution multi-view benchmark