High-res multi-view results

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

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

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

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

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

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

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

CNVR-MVS-fine		88.93 1	89.13 1	88.33 2	94.77 1	73.82 2	90.51 22	93.00 13	80.90 3	88.06 4	94.06 7	76.43 2	96.84 4	88.48 1	95.99 1	94.34 6

SteuartSystems-ACMMP		88.72 2	88.86 2	88.32 3	92.14 20	72.96 5	93.73 2	93.67 4	80.19 6	88.10 3	94.80 1	73.76 6	97.11 1	87.51 2	95.82 3	94.90 3
Skip Steuart: Steuart Systems R&D Blog.
DeepPCF-MVS	80.84 1	88.10 3	88.56 3	86.73 16	92.24 19	69.03 32	89.57 33	93.39 8	77.53 21	89.79 1	94.12 5	78.98 1	96.58 7	85.66 3	95.72 4	94.58 4

NCCC		88.06 4	88.01 4	88.24 4	94.41 3	73.62 3	91.22 16	92.83 14	81.50 1	85.79 6	93.47 11	73.02 8	97.00 3	84.90 5	94.94 9	94.10 11

APD-MVS		87.44 5	87.52 5	87.19 11	94.24 4	72.39 11	91.86 10	92.83 14	73.01 51	88.58 2	94.52 2	73.36 7	96.49 8	84.26 7	95.01 8	92.70 26

HPM-MVS		87.11 6	86.98 6	87.50 8	93.88 7	72.16 14	92.19 7	93.33 9	76.07 34	83.81 10	93.95 8	69.77 18	96.01 14	85.15 4	94.66 10	94.32 7

CP-MVS		87.11 6	86.92 7	87.68 7	94.20 5	73.86 1	93.98 1	92.82 16	76.62 26	83.68 11	94.46 3	67.93 23	95.95 15	84.20 8	94.39 12	93.23 21

DeepC-MVS	79.81 2	87.08 8	86.88 8	87.69 6	91.16 25	72.32 13	90.31 26	93.94 2	77.12 23	82.82 14	94.23 4	72.13 10	97.09 2	84.83 6	95.37 6	93.65 13
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
DeepC-MVS_fast	79.65 3	86.91 9	86.62 9	87.76 5	93.52 10	72.37 12	91.26 13	93.04 11	76.62 26	84.22 9	93.36 13	71.44 14	96.76 5	80.82 15	95.33 7	94.16 9
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
PHI-MVS		86.43 10	86.17 11	87.24 10	90.88 28	70.96 19	92.27 6	94.07 1	72.45 53	85.22 7	91.90 21	69.47 19	96.42 9	83.28 9	95.94 2	94.35 5

CSCG		86.41 11	86.19 10	87.07 13	92.91 15	72.48 10	90.81 19	93.56 5	73.95 45	83.16 13	91.07 31	75.94 3	95.19 24	79.94 18	94.38 13	93.55 16

ACMMP		85.89 12	85.39 12	87.38 9	93.59 9	72.63 8	92.74 4	93.18 10	76.78 25	80.73 23	93.82 9	64.33 42	96.29 10	82.67 10	90.69 29	93.23 21
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
CDPH-MVS		85.76 13	85.29 15	87.17 12	93.49 11	71.08 17	88.58 46	92.42 20	68.32 86	84.61 8	93.48 10	72.32 9	96.15 12	79.00 19	95.43 5	94.28 8

TSAR-MVS		85.71 14	85.33 13	86.84 14	91.34 23	72.50 9	89.07 40	87.28 87	76.41 28	85.80 5	90.22 44	74.15 5	95.37 23	81.82 11	91.88 21	92.65 28

3Dnovator+	77.84 4	85.48 15	84.47 18	88.51 1	91.08 26	73.49 4	93.18 3	93.78 3	80.79 4	76.66 63	93.37 12	60.40 73	96.75 6	77.20 31	93.73 16	95.29 1

DELS-MVS		85.41 16	85.30 14	85.77 22	88.49 64	67.93 51	85.52 86	93.44 6	78.70 15	83.63 12	89.03 59	74.57 4	95.71 17	80.26 17	94.04 14	93.66 12

HPM-MVS_fast		85.35 17	84.95 16	86.57 18	93.69 8	70.58 23	92.15 8	91.62 39	73.89 46	82.67 16	94.09 6	62.60 51	95.54 19	80.93 13	92.93 17	93.57 15

MVS_111021_HR		85.14 18	84.75 17	86.32 19	91.65 22	72.70 7	85.98 78	90.33 57	76.11 33	82.08 17	91.61 25	71.36 15	94.17 48	81.02 12	92.58 19	92.08 38

CPTT-MVS		83.73 19	83.33 20	84.92 30	93.28 13	70.86 21	92.09 9	90.38 54	68.75 83	79.57 26	92.83 15	60.60 70	93.04 68	80.92 14	91.56 25	90.86 58

EPNet		83.72 20	82.92 23	86.14 21	84.22 112	69.48 30	91.05 18	85.27 104	81.30 2	76.83 60	91.65 23	66.09 33	95.56 18	76.00 34	93.85 15	93.38 19
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
HQP_MVS		83.64 21	83.14 21	85.14 26	90.08 35	68.71 41	91.25 14	92.44 18	79.12 9	78.92 32	91.00 34	60.42 71	95.38 21	78.71 22	86.32 56	91.33 50

Vis-MVSNet		83.46 22	82.80 25	85.43 23	90.25 33	68.74 40	90.30 27	90.13 58	76.33 32	80.87 22	92.89 14	61.00 67	94.20 47	72.45 44	90.97 27	93.35 20
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
MG-MVS		83.41 23	83.45 19	83.28 55	92.74 16	62.28 102	88.17 54	89.50 61	75.22 38	81.49 18	92.74 17	66.75 29	95.11 26	72.85 41	91.58 24	92.45 29

EPP-MVSNet		83.40 24	83.02 22	84.57 35	90.13 34	64.47 81	92.32 5	90.73 51	74.45 42	79.35 28	91.10 29	69.05 20	95.12 25	72.78 42	87.22 46	94.13 10

3Dnovator	76.31 5	83.38 25	82.31 27	86.59 17	87.94 75	72.94 6	90.64 20	92.14 28	77.21 22	75.47 74	92.83 15	58.56 76	94.72 38	73.24 40	92.71 18	92.13 37

IS-MVSNet		83.15 26	82.81 24	84.18 44	89.94 37	63.30 91	91.59 11	88.46 78	79.04 11	79.49 27	92.16 18	65.10 39	94.28 44	67.71 59	91.86 22	94.95 2

DP-MVS Recon		83.11 27	82.09 29	86.15 20	94.44 2	70.92 20	88.79 43	92.20 25	70.53 66	79.17 29	91.03 33	64.12 44	96.03 13	68.39 57	90.14 33	91.50 49

PAPM_NR		83.02 28	82.41 26	84.82 32	92.47 18	66.37 64	87.93 57	91.80 35	73.82 47	77.32 52	90.66 38	67.90 24	94.90 34	70.37 48	89.48 37	93.19 23

MVSFormer		82.85 29	82.05 30	85.24 25	87.35 88	70.21 24	90.50 23	90.38 54	68.55 84	81.32 19	89.47 51	61.68 58	93.46 57	78.98 20	90.26 31	92.05 39

OMC-MVS		82.69 30	81.97 32	84.85 31	88.75 58	67.42 55	87.98 56	90.87 50	74.92 39	79.72 25	91.65 23	62.19 57	93.96 49	75.26 36	86.42 55	93.16 24

PVSNet_Blended_VisFu		82.62 31	81.83 33	84.96 29	90.80 29	69.76 29	88.74 44	91.70 38	69.39 77	78.96 31	88.46 69	65.47 36	94.87 36	74.42 38	88.57 40	90.24 68

MVS_111021_LR		82.61 32	82.11 28	84.11 45	88.82 56	71.58 15	85.15 87	86.16 99	74.69 40	80.47 24	91.04 32	62.29 55	90.55 103	80.33 16	90.08 34	90.20 69

HQP-MVS		82.61 32	82.02 31	84.37 39	89.33 44	66.98 60	89.17 36	92.19 26	76.41 28	77.23 55	90.23 43	60.17 74	95.11 26	77.47 29	85.99 60	91.03 52

CLD-MVS		82.31 34	81.65 34	84.29 41	88.47 65	67.73 54	85.81 83	92.35 23	75.78 35	78.33 40	86.58 101	64.01 45	94.35 43	76.05 33	87.48 44	90.79 59
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
LPG-MVS_test		82.08 35	81.27 36	84.50 37	89.23 50	68.76 38	90.22 28	91.94 32	75.37 36	76.64 65	91.51 26	54.29 94	94.91 32	78.44 24	83.78 67	89.83 76

API-MVS		81.99 36	81.23 37	84.26 42	90.94 27	70.18 26	91.10 17	89.32 64	71.51 60	78.66 35	88.28 73	65.26 37	95.10 29	64.74 76	91.23 26	87.51 111

UniMVSNet_NR-MVSNet		81.88 37	81.54 35	82.92 63	88.46 66	63.46 86	87.13 63	92.37 21	80.19 6	78.38 38	89.14 55	71.66 12	93.05 66	70.05 50	76.46 114	92.25 32

MAR-MVS		81.84 38	80.70 40	85.27 24	91.32 24	71.53 16	89.82 31	90.92 49	69.77 73	78.50 37	86.21 104	62.36 54	94.52 42	65.36 71	92.05 20	89.77 78
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
PAPR		81.66 39	80.89 39	83.99 48	90.27 32	64.00 84	86.76 71	91.77 37	68.84 82	77.13 59	89.50 50	67.63 25	94.88 35	67.55 60	88.52 41	93.09 25

UniMVSNet (Re)		81.60 40	81.11 38	83.09 59	88.38 69	64.41 82	87.60 59	93.02 12	78.42 18	78.56 36	88.16 74	69.78 17	93.26 60	69.58 54	76.49 113	91.60 46

ACMP	74.13 6	81.51 41	80.57 43	84.36 40	89.42 42	68.69 44	89.97 30	91.50 42	74.46 41	75.04 85	90.41 40	53.82 99	94.54 40	77.56 28	82.91 71	89.86 75
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
jason		81.39 42	80.29 47	84.70 34	86.63 94	69.90 28	85.95 79	86.77 92	63.24 122	81.07 21	89.47 51	61.08 66	92.15 83	78.33 27	90.07 35	92.05 39
jason: jason.
lupinMVS		81.39 42	80.27 48	84.76 33	87.35 88	70.21 24	85.55 85	86.41 96	62.85 126	81.32 19	88.61 66	61.68 58	92.24 82	78.41 26	90.26 31	91.83 42

CP-MVSNet		81.18 44	80.70 40	82.64 69	88.46 66	63.34 88	87.13 63	92.37 21	79.01 12	77.93 45	87.97 76	71.66 12	92.81 71	66.08 66	75.65 123	91.77 44

WR-MVS_H		81.16 45	80.66 42	82.66 68	88.05 72	63.32 90	87.38 62	92.62 17	78.53 16	77.54 51	88.08 75	70.13 16	93.00 69	67.79 58	75.81 121	91.65 45

DU-MVS		81.12 46	80.52 44	82.90 64	87.80 77	63.46 86	87.02 65	91.87 34	79.01 12	78.38 38	89.07 57	65.02 40	93.05 66	70.05 50	76.46 114	92.20 34

PVSNet_Blended		80.98 47	80.34 45	82.90 64	88.85 54	65.40 72	84.43 94	92.00 29	67.62 89	78.11 42	85.05 112	66.02 34	94.27 45	71.52 45	89.50 36	89.01 85

QAPM		80.88 48	79.50 55	85.03 27	88.01 74	68.97 35	91.59 11	92.00 29	66.63 96	75.15 83	92.16 18	57.70 78	95.45 20	63.52 77	88.76 39	90.66 61

TranMVSNet+NR-MVSNet		80.84 49	80.31 46	82.42 71	87.85 76	62.33 101	87.74 58	91.33 44	80.55 5	77.99 44	89.86 45	65.23 38	92.62 73	67.05 64	75.24 129	92.30 30

UGNet		80.83 50	79.59 53	84.54 36	88.04 73	68.09 50	89.42 34	88.16 79	76.95 24	76.22 68	89.46 53	49.30 118	93.94 51	68.48 56	90.31 30	91.60 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
XVG-OURS-SEG-HR		80.81 51	79.76 52	83.96 50	85.60 102	68.78 37	83.54 103	90.50 52	70.66 65	76.71 62	91.66 22	60.69 69	91.26 90	76.94 32	81.58 83	91.83 42

DTE-MVSNet		80.80 52	80.26 49	82.41 72	87.68 81	62.36 99	86.78 68	91.09 46	80.19 6	77.82 47	89.14 55	67.20 27	92.61 74	67.14 62	75.33 126	92.25 32

ACMM	73.20 8	80.78 53	79.84 51	83.58 51	89.31 47	68.37 47	89.99 29	91.60 40	70.28 69	77.25 53	89.66 48	53.37 102	93.53 56	74.24 39	82.85 72	88.85 88
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
114514_t		80.68 54	79.51 54	84.20 43	94.09 6	67.27 59	89.64 32	91.11 45	58.75 147	74.08 90	90.72 37	58.10 77	95.04 31	69.70 53	89.42 38	90.30 67

PVSNet_BlendedMVS		80.60 55	80.02 50	82.36 74	88.85 54	65.40 72	86.16 76	92.00 29	69.34 78	78.11 42	86.09 105	66.02 34	94.27 45	71.52 45	82.06 78	87.39 113

AdaColmap		80.58 56	79.42 56	84.06 47	93.09 14	68.91 36	89.36 35	88.97 74	69.27 79	75.70 73	89.69 47	57.20 81	95.77 16	63.06 80	88.41 42	87.50 112

XVG-OURS		80.41 57	79.23 60	83.97 49	85.64 101	69.02 33	83.03 106	90.39 53	71.09 62	77.63 50	91.49 28	54.62 93	91.35 88	75.71 35	83.47 69	91.54 48

PCF-MVS	73.52 7	80.38 58	78.84 64	85.01 28	87.71 80	68.99 34	83.65 100	91.46 43	63.00 124	77.77 49	90.28 41	66.10 32	95.09 30	61.40 92	88.22 43	90.94 57
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
test_djsdf		80.30 59	79.32 59	83.27 56	83.98 113	65.37 74	90.50 23	90.38 54	68.55 84	76.19 69	88.70 64	56.44 83	93.46 57	78.98 20	80.14 94	90.97 56

NR-MVSNet		80.23 60	79.38 57	82.78 67	87.80 77	63.34 88	86.31 74	91.09 46	79.01 12	72.17 101	89.07 57	67.20 27	92.81 71	66.08 66	75.65 123	92.20 34

IterMVS-LS		80.06 61	79.38 57	82.11 75	85.89 97	63.20 93	86.79 67	89.34 63	74.19 43	75.45 75	86.72 95	66.62 30	92.39 80	72.58 43	76.86 111	90.75 60
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
OpenMVS	72.83 10	79.77 62	78.33 69	84.09 46	85.17 104	69.91 27	90.57 21	90.97 48	66.70 95	72.17 101	91.91 20	54.70 91	93.96 49	61.81 90	90.95 28	88.41 97

BH-RMVSNet		79.61 63	78.44 68	83.14 58	89.38 43	65.93 67	84.95 89	87.15 88	73.56 50	78.19 41	89.79 46	56.67 82	93.36 59	59.53 106	86.74 52	90.13 70

WR-MVS		79.51 64	79.04 62	80.90 90	87.47 87	59.41 119	88.85 42	93.40 7	77.81 19	73.66 91	88.97 60	67.25 26	92.27 81	60.28 101	73.71 137	91.01 55

ab-mvs		79.51 64	78.97 63	81.14 88	88.46 66	60.91 107	83.84 98	89.24 68	70.36 68	79.03 30	88.87 62	63.23 48	90.21 105	65.12 72	82.57 76	92.28 31

BH-untuned		79.47 66	78.60 66	82.05 76	89.19 52	65.91 68	86.07 77	88.52 77	72.18 54	75.42 76	87.69 80	61.15 65	93.54 55	60.38 99	86.83 51	86.70 126

mvs_anonymous		79.42 67	79.11 61	80.34 96	84.45 111	57.97 126	82.59 107	87.62 84	67.40 93	76.17 71	88.56 68	68.47 21	89.59 112	70.65 47	86.05 59	93.47 18

TAPA-MVS	73.13 9	79.15 68	77.94 72	82.79 66	89.59 41	62.99 97	88.16 55	91.51 41	65.77 102	77.14 58	91.09 30	60.91 68	93.21 61	50.26 141	87.05 49	92.17 36
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
CDS-MVSNet		79.07 69	77.70 74	83.17 57	87.60 82	68.23 49	84.40 95	86.20 98	67.49 91	76.36 67	86.54 102	61.54 60	90.79 102	61.86 89	87.33 45	90.49 63
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
MVSTER		79.01 70	77.88 73	82.38 73	83.07 122	64.80 79	84.08 97	88.95 75	69.01 81	78.69 34	87.17 92	54.70 91	92.43 79	74.69 37	80.57 90	89.89 74

TAMVS		78.89 71	77.51 75	83.03 61	87.80 77	67.79 53	84.72 92	85.05 107	67.63 88	76.75 61	87.70 79	62.25 56	90.82 101	58.53 111	87.13 48	90.49 63

diffmvs		78.78 72	78.64 65	79.22 107	84.48 110	56.11 147	82.12 113	87.53 86	65.53 105	76.66 63	88.73 63	68.25 22	87.97 121	68.73 55	85.70 63	93.59 14

MVS_Test		78.49 73	78.26 71	79.18 111	83.29 118	56.48 140	82.45 108	86.59 94	66.73 94	76.13 72	88.46 69	66.33 31	89.04 116	70.30 49	84.06 66	92.01 41

GBi-Net		78.40 74	77.40 76	81.40 83	87.60 82	63.01 94	88.39 49	89.28 65	71.63 57	75.34 78	87.28 87	54.80 88	91.11 93	62.72 81	79.57 95	90.09 71

test1		78.40 74	77.40 76	81.40 83	87.60 82	63.01 94	88.39 49	89.28 65	71.63 57	75.34 78	87.28 87	54.80 88	91.11 93	62.72 81	79.57 95	90.09 71

Vis-MVSNet (Re-imp)		78.36 76	78.45 67	78.07 119	88.64 60	51.78 157	86.70 72	79.63 142	74.14 44	75.11 84	90.83 36	61.29 64	89.75 111	58.10 115	91.60 23	92.69 27

BH-w/o		78.21 77	77.33 78	80.84 91	88.81 57	65.13 78	84.87 90	87.85 81	69.75 74	74.52 88	84.74 114	61.34 62	93.11 65	58.24 114	85.84 62	84.27 139

FMVSNet2		78.20 78	77.21 79	81.20 86	87.60 82	62.89 98	87.47 61	89.02 72	71.63 57	75.29 82	87.28 87	54.80 88	91.10 96	62.38 85	79.38 98	89.61 80

Baseline_NR-MVSNet		78.15 79	78.33 69	77.61 122	85.79 98	56.21 145	86.78 68	85.76 102	73.60 48	77.93 45	87.57 83	65.02 40	88.99 117	67.14 62	75.33 126	87.63 108

CNLPA		78.08 80	76.79 82	81.97 78	90.40 31	71.07 18	87.59 60	84.55 108	66.03 101	72.38 100	89.64 49	57.56 79	86.04 132	59.61 104	83.35 70	88.79 89

PLC	70.83 11	78.05 81	76.37 85	83.08 60	91.88 21	67.80 52	88.19 53	89.46 62	64.33 116	69.87 120	88.38 71	53.66 101	93.58 54	58.86 109	82.73 74	87.86 105
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
HY-MVS	69.67 12	77.95 82	77.15 80	80.36 95	87.57 86	60.21 112	83.37 105	87.78 82	66.11 99	75.37 77	87.06 94	63.27 46	90.48 104	61.38 93	82.43 77	90.40 66

FMVSNet3		77.88 83	76.85 81	80.97 89	86.84 92	62.36 99	86.52 73	88.77 76	71.13 61	75.34 78	86.66 100	54.07 97	91.10 96	62.72 81	79.57 95	89.45 81

PAPM		77.68 84	76.40 84	81.51 81	87.29 90	61.85 104	83.78 99	89.59 60	64.74 111	71.23 109	88.70 64	62.59 52	93.66 53	52.66 135	87.03 50	89.01 85

FMVSNet1		77.44 85	76.12 87	81.40 83	86.81 93	63.01 94	88.39 49	89.28 65	70.49 67	74.39 89	87.28 87	49.06 120	91.11 93	60.91 96	78.52 100	90.09 71

TR-MVS		77.44 85	76.18 86	81.20 86	88.24 70	63.24 92	84.61 93	86.40 97	67.55 90	77.81 48	86.48 103	54.10 96	93.15 64	57.75 116	82.72 75	87.20 116

1112_ss		77.40 87	76.43 83	80.32 97	89.11 53	60.41 111	83.65 100	87.72 83	62.13 131	73.05 93	86.72 95	62.58 53	89.97 109	62.11 87	80.80 88	90.59 62

LS3D		76.95 88	74.82 92	83.37 53	90.45 30	67.36 58	89.15 39	86.94 91	61.87 133	69.52 122	90.61 39	51.71 110	94.53 41	46.38 154	86.71 53	88.21 100

DP-MVS		76.78 89	74.57 94	83.42 52	93.29 12	69.46 31	88.55 47	83.70 114	63.98 120	70.20 115	88.89 61	54.01 98	94.80 37	46.66 151	81.88 81	86.01 130

cascas		76.72 90	74.64 93	82.99 62	85.78 100	65.88 69	82.33 111	89.21 69	60.85 137	72.74 95	81.02 136	47.28 122	93.75 52	67.48 61	85.02 64	89.34 82

Test_1112_low_res		76.40 91	75.44 90	79.27 105	89.28 48	58.09 125	81.69 115	87.07 89	59.53 144	72.48 99	86.67 99	61.30 63	89.33 114	60.81 98	80.15 93	90.41 65

F-COLMAP		76.38 92	74.33 97	82.50 70	89.28 48	66.95 62	88.41 48	89.03 71	64.05 118	66.83 134	88.61 66	46.78 124	92.89 70	57.48 117	78.55 99	87.67 107

LTVRE_ROB	69.57 13	76.25 93	74.54 96	81.41 82	88.60 61	64.38 83	79.24 130	89.12 70	70.76 64	69.79 121	87.86 77	49.09 119	93.20 62	56.21 124	80.16 92	86.65 127
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
XVG-ACMP-BASELINE		76.11 94	74.27 98	81.62 80	83.20 119	64.67 80	83.60 102	89.75 59	69.75 74	71.85 104	87.09 93	32.78 160	92.11 84	69.99 52	80.43 91	88.09 102

ACMH+	68.96 14	76.01 95	74.01 99	82.03 77	88.60 61	65.31 75	88.86 41	87.55 85	70.25 70	67.75 129	87.47 85	41.27 144	93.19 63	58.37 112	75.94 119	87.60 110

ACMH	67.68 16	75.89 96	73.93 100	81.77 79	88.71 59	66.61 63	88.62 45	89.01 73	69.81 72	66.78 135	86.70 98	41.95 143	91.51 87	55.64 125	78.14 104	87.17 117
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
IB-MVS	68.01 15	75.85 97	73.36 105	83.31 54	84.76 107	66.03 65	83.38 104	85.06 106	70.21 71	69.40 123	81.05 135	45.76 129	94.66 39	65.10 73	75.49 125	89.25 83
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
WTY-MVS		75.65 98	75.68 88	75.57 134	86.40 95	56.82 135	77.92 136	82.40 124	65.10 110	76.18 70	87.72 78	63.13 49	80.90 149	60.31 100	81.96 79	89.00 87

EPNet_dtu		75.46 99	74.86 91	77.23 125	82.57 131	54.60 151	86.89 66	83.09 117	71.64 56	66.25 137	85.86 107	55.99 85	88.04 120	54.92 126	86.55 54	89.05 84
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
XXY-MVS		75.41 100	75.56 89	74.96 136	83.59 115	57.82 129	80.59 123	83.87 113	66.54 97	74.93 86	88.31 72	63.24 47	80.09 152	62.16 86	76.85 112	86.97 122

TransMVSNet (Re)		75.39 101	74.56 95	77.86 120	85.50 103	57.10 134	86.78 68	86.09 101	72.17 55	71.53 108	87.34 86	63.01 50	89.31 115	56.84 121	61.83 157	87.17 117

CostFormer		75.24 102	73.90 101	79.27 105	82.65 130	58.27 124	80.80 119	82.73 121	61.57 134	75.33 81	83.13 120	55.52 86	91.07 99	64.98 75	78.34 103	88.45 95

PT_06_test		75.20 103	73.77 103	79.49 104	82.69 129	60.19 113	82.34 109	86.99 90	69.71 76	72.52 98	78.31 147	56.27 84	90.07 108	62.03 88	73.11 138	88.23 98

PEN-MVS		74.79 104	73.85 102	77.61 122	85.79 98	56.21 145	86.31 74	85.76 102	73.60 48	72.17 101	87.57 83	61.36 61	88.99 117	58.98 108	53.14 167	87.63 108

PatchFormer-LS_test		74.50 105	73.05 106	78.86 113	82.95 125	59.55 118	81.65 116	82.30 126	67.44 92	71.62 107	78.15 150	52.34 105	88.92 119	65.05 74	75.90 120	88.12 101

IterMVS		74.29 106	72.94 107	78.35 117	81.53 139	63.49 85	81.58 117	82.49 122	68.06 87	69.99 119	83.69 117	51.66 111	85.54 134	65.85 68	71.64 143	86.01 130
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
OurMVSNet-221017-0		74.26 107	72.42 108	79.80 100	83.76 114	59.59 116	85.92 81	86.64 93	66.39 98	66.96 133	87.58 82	39.46 148	91.60 86	65.76 69	69.27 147	88.22 99

EG-PatchMatch MVS		74.04 108	71.82 112	80.71 93	84.92 106	67.42 55	85.86 82	88.08 80	66.04 100	64.22 145	83.85 115	35.10 159	92.56 77	57.44 118	80.83 87	82.16 149

Test4		73.95 109	72.20 109	79.21 108	82.91 126	58.94 121	81.25 118	82.17 127	65.21 107	71.05 111	82.44 124	44.21 134	90.17 106	63.29 78	77.28 108	88.53 93

PatchFormer_test		73.70 110	71.86 111	79.21 108	82.91 126	58.94 121	82.34 109	82.17 127	65.21 107	71.05 111	78.31 147	44.21 134	90.17 106	63.29 78	77.28 108	88.53 93

sss		73.60 111	73.64 104	73.51 143	82.80 128	55.01 150	76.12 140	81.69 131	62.47 130	74.68 87	85.85 108	57.32 80	78.11 156	60.86 97	80.93 86	87.39 113

tpmp4_e23		73.45 112	71.17 119	80.31 98	83.55 116	59.56 117	81.88 114	82.33 125	57.94 151	70.51 114	81.62 131	51.19 114	91.63 85	53.96 130	77.51 107	89.75 79

SixPastTwentyTwo		73.37 113	71.26 118	79.70 101	85.08 105	57.89 128	85.57 84	83.56 115	71.03 63	65.66 139	85.88 106	42.10 141	92.57 76	59.11 107	63.34 155	88.65 92

CR-MVSNet		73.37 113	71.27 117	79.67 102	81.32 140	65.19 76	75.92 141	80.30 138	59.92 141	72.73 96	81.19 133	52.50 103	86.69 127	59.84 102	77.71 105	87.11 120

MSDG		73.36 115	70.99 120	80.49 94	84.51 109	65.80 70	80.71 121	86.13 100	65.70 103	65.46 140	83.74 116	44.60 132	90.91 100	51.13 139	76.89 110	84.74 137

tpm2		73.26 116	71.46 114	78.63 114	83.34 117	56.71 137	80.65 122	80.40 137	56.63 155	73.55 92	82.02 129	51.80 109	91.24 91	56.35 123	78.42 102	87.95 103

RPSCF		73.23 117	71.46 114	78.54 115	82.50 132	59.85 115	82.18 112	82.84 120	58.96 145	71.15 110	89.41 54	45.48 131	84.77 136	58.82 110	71.83 142	91.02 54

PatchmatchNet		73.12 118	71.33 116	78.49 116	83.18 120	60.85 108	79.63 127	78.57 145	64.13 117	71.73 105	79.81 144	51.20 113	85.97 133	57.40 119	76.36 116	88.66 91
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
COLMAP_ROB	66.92 17	73.01 119	70.41 125	80.81 92	87.13 91	65.63 71	88.30 52	84.19 111	62.96 125	63.80 147	87.69 80	38.04 153	92.56 77	46.66 151	74.91 130	84.24 140
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
test_0402		72.79 120	70.44 124	79.84 99	88.13 71	65.99 66	85.93 80	84.29 110	65.57 104	67.40 131	85.49 109	46.92 123	92.61 74	35.88 165	74.38 133	80.94 151

tpmrst		72.39 121	72.13 110	73.18 145	80.54 146	49.91 163	79.91 126	79.08 144	63.11 123	71.69 106	79.95 141	55.32 87	82.77 144	65.66 70	73.89 136	86.87 123

PatchMatch-RL		72.38 122	70.90 121	76.80 128	88.60 61	67.38 57	79.53 128	76.17 152	62.75 128	69.36 124	82.00 130	45.51 130	84.89 135	53.62 131	80.58 89	78.12 159

tpm		72.37 123	71.71 113	74.35 139	82.19 134	52.00 155	79.22 131	77.29 149	64.56 113	72.95 94	83.68 118	51.35 112	83.26 141	58.33 113	75.80 122	87.81 106

PVSNet	64.34 18	72.08 124	70.87 122	75.69 133	86.21 96	56.44 142	74.37 149	80.73 135	62.06 132	70.17 116	82.23 127	42.86 139	83.31 140	54.77 127	84.45 65	87.32 115

RPMNet		71.62 125	68.94 130	79.67 102	81.32 140	65.19 76	75.92 141	78.30 148	57.60 152	72.73 96	76.45 157	52.30 106	86.69 127	48.14 148	77.71 105	87.11 120

HyFIR		71.37 126	70.50 123	73.97 142	80.52 147	59.87 114	70.92 155	82.42 123	56.28 156	65.84 138	76.50 156	53.72 100	83.21 143	61.07 95	87.21 47	78.84 157

K. test v3		71.19 127	68.51 132	79.21 108	83.04 123	57.78 130	84.35 96	76.91 151	72.90 52	62.99 149	82.86 121	39.27 149	91.09 98	61.65 91	52.66 168	88.75 90

tpmvs		71.09 128	69.29 128	76.49 129	82.04 135	56.04 148	78.92 134	81.37 133	64.05 118	67.18 132	78.28 149	49.74 116	89.77 110	49.67 144	72.37 139	83.67 141

Patchmtry		70.74 129	69.16 129	75.49 135	80.72 143	54.07 152	74.94 148	80.30 138	58.34 148	70.01 117	81.19 133	52.50 103	86.54 129	53.37 132	71.09 144	85.87 132

MIMVSNet		70.69 130	69.30 127	74.88 137	84.52 108	56.35 143	75.87 143	79.42 143	64.59 112	67.76 128	82.41 125	41.10 145	81.54 148	46.64 153	81.34 84	86.75 125

tpm cat1		70.57 131	68.31 134	77.35 124	82.41 133	57.95 127	78.08 135	80.22 140	52.04 161	68.54 127	77.66 153	52.00 108	87.84 123	51.77 136	72.07 141	86.25 129

OpenMVS_ROB	64.09 19	70.56 132	68.19 135	77.65 121	80.26 148	59.41 119	85.01 88	82.96 119	58.76 146	65.43 141	82.33 126	37.63 155	91.23 92	45.34 157	76.03 118	82.32 147

USDC		70.33 133	68.37 133	76.21 131	80.60 145	56.23 144	79.19 132	86.49 95	60.89 136	61.29 150	85.47 110	31.78 161	89.47 113	53.37 132	76.21 117	82.94 146

CMPMVS	51.72 21	70.19 134	68.16 136	76.28 130	73.15 166	57.55 131	79.47 129	83.92 112	48.02 166	56.48 160	84.81 113	43.13 137	86.42 131	62.67 84	81.81 82	84.89 136
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
FMVSNet5		69.50 135	67.96 138	74.15 140	82.97 124	55.35 149	80.01 125	82.12 129	62.56 129	63.02 148	81.53 132	36.92 156	81.92 146	48.42 147	74.06 134	85.17 135

PMMVS		69.34 136	68.67 131	71.35 152	75.67 160	62.03 103	75.17 144	73.46 160	50.00 164	68.68 126	79.05 145	52.07 107	78.13 155	61.16 94	82.77 73	73.90 162

EPMVS		69.02 137	68.16 136	71.59 148	79.61 152	49.80 164	77.40 137	66.93 167	62.82 127	70.01 117	79.05 145	45.79 128	77.86 157	56.58 122	75.26 128	87.13 119

MIMVSNet1		68.58 138	66.78 140	73.98 141	80.07 149	51.82 156	80.77 120	84.37 109	64.40 115	59.75 153	82.16 128	36.47 157	83.63 138	42.73 159	70.33 145	86.48 128

PatchT		68.46 139	67.85 139	70.29 154	80.70 144	43.93 169	72.47 152	74.88 156	60.15 140	70.55 113	76.57 155	49.94 115	81.59 147	50.58 140	74.83 131	85.34 134

TDRefinement		67.49 140	64.34 144	76.92 126	73.47 165	61.07 106	84.86 91	82.98 118	59.77 142	58.30 156	85.13 111	26.06 163	87.89 122	47.92 149	60.59 160	81.81 150

UnsupCasMVSNet_eth		67.33 141	65.99 141	71.37 150	73.48 164	51.47 159	75.16 145	85.19 105	65.20 109	60.78 152	80.93 138	42.35 140	77.20 159	57.12 120	53.69 166	85.44 133

TinyColmap		67.30 142	64.81 143	74.76 138	81.92 136	56.68 138	80.29 124	81.49 132	60.33 139	56.27 161	83.22 119	24.77 164	87.66 124	45.52 156	69.47 146	79.95 154

dp		66.80 143	65.43 142	70.90 153	79.74 150	48.82 167	75.12 147	74.77 157	59.61 143	64.08 146	77.23 154	42.89 138	80.72 150	48.86 146	66.58 150	83.16 142

MDA-MVSNet-bldmvs		66.68 144	63.66 146	75.75 132	79.28 154	60.56 110	73.92 150	78.35 147	64.43 114	50.13 167	79.87 143	44.02 136	83.67 137	46.10 155	56.86 162	83.03 144

PM-MVS		66.41 145	64.14 145	73.20 144	73.92 162	56.45 141	78.97 133	64.96 170	63.88 121	64.72 143	80.24 140	19.84 168	83.44 139	66.24 65	64.52 153	79.71 155

JIA-IIPM		66.32 146	62.82 150	76.82 127	77.09 156	61.72 105	65.34 163	75.38 154	58.04 150	64.51 144	62.32 166	42.05 142	86.51 130	51.45 138	69.22 148	82.21 148

PS-CasMVS		65.34 147	63.02 147	72.29 146	79.74 150	49.05 166	64.60 165	78.57 145	65.26 106	65.05 142	82.68 123	45.86 126	83.22 142	43.19 158	42.23 171	84.61 138

YYNet1		65.03 148	62.91 149	71.38 149	75.85 159	56.60 139	69.12 159	74.66 159	57.28 153	54.12 163	77.87 152	45.85 127	74.48 163	49.95 142	61.52 159	83.05 143

MDA-MVSNet_test_wron		65.03 148	62.92 148	71.37 150	75.93 158	56.73 136	69.09 160	74.73 158	57.28 153	54.03 164	77.89 151	45.88 125	74.39 164	49.89 143	61.55 158	82.99 145

LF4IMVS		64.02 150	62.19 151	69.50 156	70.90 168	53.29 154	76.13 139	77.18 150	52.65 160	58.59 154	80.98 137	23.55 165	76.52 160	53.06 134	66.66 149	78.68 158

UnsupCasMVSNet_bld		63.70 151	61.53 152	70.21 155	73.69 163	51.39 160	72.82 151	81.89 130	55.63 158	57.81 157	71.80 160	38.67 151	78.61 154	49.26 145	52.21 169	80.63 152

PVSNet_0	57.27 20	61.67 152	59.27 153	68.85 157	79.61 152	57.44 132	68.01 161	73.44 161	55.93 157	58.54 155	70.41 161	44.58 133	77.55 158	47.01 150	35.91 172	71.55 163

LP		61.36 153	57.78 154	72.09 147	75.54 161	58.53 123	67.16 162	75.22 155	51.90 162	54.13 162	69.97 162	37.73 154	80.45 151	32.74 168	55.63 164	77.29 160

MVS-HIRNet		59.14 154	57.67 155	63.57 160	81.65 138	43.50 170	71.73 153	65.06 169	39.59 169	51.43 166	57.73 168	38.34 152	82.58 145	39.53 161	73.95 135	64.62 167

DSMNet-mixed		57.77 155	56.90 157	60.38 161	67.70 171	35.61 175	69.18 158	53.97 173	32.30 175	57.49 158	79.88 142	40.39 147	68.57 170	38.78 162	72.37 139	76.97 161

testpf		56.51 156	57.58 156	53.30 167	71.99 167	41.19 172	46.89 172	69.32 165	58.06 149	52.87 165	69.45 163	27.99 162	72.73 166	59.59 105	62.07 156	45.98 173

FPMVS		53.68 157	51.64 159	59.81 162	65.08 172	51.03 161	69.48 157	69.58 163	41.46 167	40.67 170	72.32 159	16.46 171	70.00 169	24.24 172	65.42 151	58.40 169

N_pmnet		52.79 158	53.26 158	51.40 169	78.99 155	7.68 182	69.52 156	3.89 179	51.63 163	57.01 159	74.98 158	40.83 146	65.96 171	37.78 163	64.67 152	80.56 153

HyFIR lowres test		51.79 159	50.01 161	57.11 164	68.82 169	49.21 165	60.50 167	53.26 174	34.52 171	43.77 169	64.94 165	20.34 167	71.75 167	39.87 160	64.06 154	50.39 170

no-one		51.08 160	45.79 163	66.95 158	57.92 175	50.49 162	59.63 169	76.04 153	48.04 165	31.85 171	56.10 171	19.12 169	80.08 153	36.89 164	26.52 174	70.29 164

new_pmnet		50.91 161	50.29 160	52.78 168	68.58 170	34.94 178	63.71 166	56.63 172	39.73 168	44.95 168	65.47 164	21.93 166	58.48 173	34.98 166	56.62 163	64.92 166

ANet_high		50.57 162	46.10 162	63.99 159	48.67 177	39.13 173	70.99 154	80.85 134	61.39 135	31.18 173	57.70 169	17.02 170	73.65 165	31.22 169	15.89 179	79.18 156

Gipuma		45.18 163	41.86 164	55.16 166	77.03 157	51.52 158	32.50 174	80.52 136	32.46 173	27.12 174	35.02 175	9.52 173	75.50 162	22.31 173	60.21 161	38.45 175
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PMVS	37.38 22	44.16 164	40.28 165	55.82 165	40.82 180	42.54 171	65.12 164	63.99 171	34.43 172	24.48 175	57.12 170	3.92 175	76.17 161	17.10 175	55.52 165	48.75 171
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
PMMVS2		40.82 165	38.86 166	46.69 171	53.84 176	16.45 181	48.61 171	49.92 176	37.49 170	31.67 172	60.97 167	8.14 174	56.42 174	28.42 170	30.72 173	67.19 165

wuykxyi23d		39.76 166	33.18 168	59.51 163	46.98 178	44.01 168	57.70 170	67.74 166	24.13 176	13.98 179	34.33 176	1.27 178	71.33 168	34.23 167	18.23 177	63.18 168

PNet_i23d		38.26 167	35.42 167	46.79 170	58.74 173	35.48 176	59.65 168	51.25 175	32.45 174	23.44 177	47.53 173	2.04 177	58.96 172	25.60 171	18.09 178	45.92 174

MVE	26.22 23	30.37 168	25.89 169	43.81 172	44.55 179	35.46 177	28.87 175	39.07 177	18.20 177	18.58 178	40.18 174	2.68 176	47.37 175	17.07 176	23.78 176	48.60 172
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
wuyk23d		16.82 169	15.94 170	19.46 174	58.74 173	31.45 179	39.22 173	3.74 180	6.84 179	6.04 180	2.70 177	1.27 178	24.29 177	10.54 177	14.40 180	2.63 177

ab-mvs-re		7.23 170	9.64 171	0.00 175	0.00 182	0.00 183	0.00 176	0.00 181	0.00 180	0.00 181	86.72 95	0.00 180	0.00 178	0.00 178	0.00 181	0.00 178

test12				86.80 15	92.63 17	70.70 22		91.79 36		82.71 15		71.67 11	96.16 11		94.50 11	93.54 17

plane_prior7					90.08 35	68.51 46

plane_prior6					89.84 39	68.70 43						60.42 71

plane_prior5								92.44 18					95.38 21	78.71 22	86.32 56	91.33 50

plane_prior4											91.00 34

plane_prior3						68.60 45			78.44 17	78.92 32

plane_prior2							91.25 14		79.12 9

plane_prior1					89.90 38

plane_prior						68.71 41	90.38 25		77.62 20						86.16 58

n2								0.00 181

nn								0.00 181

door-mid								69.98 162

lessismore_v0				78.97 112	81.01 142	57.15 133		65.99 168		61.16 151	82.82 122	39.12 150	91.34 89	59.67 103	46.92 170	88.43 96

LGP-MVS_train				84.50 37	89.23 50	68.76 38		91.94 32	75.37 36	76.64 65	91.51 26	54.29 94	94.91 32	78.44 24	83.78 67	89.83 76

test11								92.23 24

door								69.44 164

HQP5-MVS						66.98 60

HQP-NCC					89.33 44		89.17 36		76.41 28	77.23 55

ACMP_Plane					89.33 44		89.17 36		76.41 28	77.23 55

BP-MVS														77.47 29

HQP4-MVS										77.24 54			95.11 26			91.03 52

HQP3-MVS								92.19 26							85.99 60

HQP2-MVS												60.17 74

NP-MVS					89.62 40	68.32 48					90.24 42

MDTV_nov1_ep13_2view						37.79 174	75.16 145		55.10 159	66.53 136		49.34 117		53.98 129		87.94 104

MDTV_nov1_ep13			69.97 126		83.18 120	53.48 153	77.10 138	80.18 141	60.45 138	69.33 125	80.44 139	48.89 121	86.90 126	51.60 137	78.51 101

ACMMP++_ref															81.95 80

ACMMP++															81.25 85

Test By Simon												64.33 42

ITE_SJBPF				78.22 118	81.77 137	60.57 109		83.30 116	69.25 80	67.54 130	87.20 91	36.33 158	87.28 125	54.34 128	74.62 132	86.80 124

DeepMVS_CX				27.40 173	40.17 181	26.90 180		24.59 178	17.44 178	23.95 176	48.61 172	9.77 172	26.48 176	18.06 174	24.47 175	28.83 176

High-resolution multi-view benchmark