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.

SteuartSystems-ACMMP		88.72 1	88.86 1	88.32 2	92.14 17	72.96 4	93.73 2	93.67 4	80.19 4	88.10 3	94.80 1	73.76 4	97.11 1	87.51 1	95.82 2	94.90 3
Skip Steuart: Steuart Systems R&D Blog.
DeepPCF-MVS	80.84 1	88.10 2	88.56 2	86.73 12	92.24 16	69.03 28	89.57 29	93.39 7	77.53 7	89.79 1	94.12 5	78.98 1	96.58 6	85.66 2	95.72 3	94.58 4

NCCC		88.06 3	88.01 3	88.24 3	94.41 2	73.62 2	91.22 14	92.83 11	81.50 1	85.79 4	93.47 9	73.02 6	97.00 3	84.90 4	94.94 7	94.10 9

APD-MVS		87.44 4	87.52 4	87.19 10	94.24 3	72.39 9	91.86 10	92.83 11	73.01 33	88.58 2	94.52 2	73.36 5	96.49 7	84.26 6	95.01 6	92.70 21

HPM-MVS		87.11 5	86.98 5	87.50 7	93.88 6	72.16 12	92.19 7	93.33 8	76.07 18	83.81 7	93.95 7	69.77 10	96.01 11	85.15 3	94.66 8	94.32 6

CP-MVS		87.11 5	86.92 6	87.68 6	94.20 4	73.86 1	93.98 1	92.82 13	76.62 12	83.68 8	94.46 3	67.93 13	95.95 12	84.20 7	94.39 9	93.23 16

DeepC-MVS	79.81 2	87.08 7	86.88 7	87.69 5	91.16 21	72.32 11	90.31 22	93.94 2	77.12 9	82.82 11	94.23 4	72.13 7	97.09 2	84.83 5	95.37 4	93.65 11
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 8	86.62 8	87.76 4	93.52 9	72.37 10	91.26 13	93.04 10	76.62 12	84.22 6	93.36 11	71.44 8	96.76 4	80.82 13	95.33 5	94.16 7
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
CSCG		86.41 10	86.19 9	87.07 11	92.91 13	72.48 8	90.81 17	93.56 5	73.95 29	83.16 10	91.07 29	75.94 2	95.19 18	79.94 16	94.38 10	93.55 13

PHI-MVS		86.43 9	86.17 10	87.24 9	90.88 24	70.96 16	92.27 6	94.07 1	72.45 35	85.22 5	91.90 19	69.47 11	96.42 8	83.28 8	95.94 1	94.35 5

ACMMP		85.89 11	85.39 11	87.38 8	93.59 8	72.63 7	92.74 4	93.18 9	76.78 11	80.73 19	93.82 8	64.33 25	96.29 9	82.67 9	90.69 25	93.23 16
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
DELS-MVS		85.41 13	85.30 12	85.77 18	88.49 54	67.93 41	85.52 67	93.44 6	78.70 6	83.63 9	89.03 48	74.57 3	95.71 14	80.26 15	94.04 11	93.66 10

HPM-MVS_fast		85.35 14	84.95 13	86.57 14	93.69 7	70.58 19	92.15 8	91.62 26	73.89 30	82.67 12	94.09 6	62.60 33	95.54 16	80.93 11	92.93 14	93.57 12

MVS_111021_HR		85.14 15	84.75 14	86.32 15	91.65 19	72.70 6	85.98 59	90.33 41	76.11 17	82.08 13	91.61 23	71.36 9	94.17 41	81.02 10	92.58 16	92.08 27

3Dnovator+	77.84 4	85.48 12	84.47 15	88.51 1	91.08 22	73.49 3	93.18 3	93.78 3	80.79 3	76.66 48	93.37 10	60.40 52	96.75 5	77.20 25	93.73 13	95.29 1

MG-MVS		83.41 19	83.45 16	83.28 49	92.74 14	62.28 83	88.17 46	89.50 45	75.22 22	81.49 14	92.74 15	66.75 16	95.11 20	72.85 35	91.58 20	92.45 23

CPTT-MVS		83.73 16	83.33 17	84.92 25	93.28 11	70.86 18	92.09 9	90.38 38	68.75 64	79.57 22	92.83 13	60.60 51	93.04 58	80.92 12	91.56 21	90.86 39

EPP-MVSNet		83.40 20	83.02 18	84.57 30	90.13 30	64.47 70	92.32 5	90.73 35	74.45 26	79.35 24	91.10 27	69.05 12	95.12 19	72.78 36	87.22 42	94.13 8

EPNet		83.72 17	82.92 19	86.14 17	84.22 88	69.48 26	91.05 16	85.27 81	81.30 2	76.83 45	91.65 21	66.09 19	95.56 15	76.00 28	93.85 12	93.38 14

IS-MVSNet		83.15 22	82.81 20	84.18 38	89.94 31	63.30 74	91.59 11	88.46 62	79.04 5	79.49 23	92.16 16	65.10 24	94.28 37	67.71 46	91.86 18	94.95 2

Vis-MVSNet		83.46 18	82.80 21	85.43 19	90.25 29	68.74 36	90.30 23	90.13 42	76.33 16	80.87 18	92.89 12	61.00 48	94.20 40	72.45 38	90.97 23	93.35 15
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
PAPM_NR		83.02 24	82.41 22	84.82 27	92.47 15	66.37 53	87.93 49	91.80 23	73.82 31	77.32 38	90.66 34	67.90 14	94.90 27	70.37 41	89.48 33	93.19 18

3Dnovator	76.31 5	83.38 21	82.31 23	86.59 13	87.94 61	72.94 5	90.64 18	92.14 17	77.21 8	75.47 56	92.83 13	58.56 54	94.72 31	73.24 34	92.71 15	92.13 26

MVS_111021_LR		82.61 28	82.11 24	84.11 39	88.82 46	71.58 13	85.15 68	86.16 79	74.69 24	80.47 20	91.04 30	62.29 37	90.55 87	80.33 14	90.08 30	90.20 50

DP-MVS Recon		83.11 23	82.09 25	86.15 16	94.44 1	70.92 17	88.79 36	92.20 15	70.53 47	79.17 25	91.03 31	64.12 27	96.03 10	68.39 45	90.14 29	91.50 34

MVSFormer		82.85 25	82.05 26	85.24 21	87.35 69	70.21 20	90.50 20	90.38 38	68.55 65	81.32 15	89.47 44	61.68 40	93.46 50	78.98 17	90.26 27	92.05 28

OMC-MVS		82.69 26	81.97 27	84.85 26	88.75 48	67.42 45	87.98 48	90.87 34	74.92 23	79.72 21	91.65 21	62.19 39	93.96 42	75.26 30	86.42 51	93.16 19

PVSNet_Blended_VisFu		82.62 27	81.83 28	84.96 24	90.80 25	69.76 25	88.74 37	91.70 25	69.39 58	78.96 27	88.46 55	65.47 22	94.87 29	74.42 32	88.57 36	90.24 49

CLD-MVS		82.31 29	81.65 29	84.29 35	88.47 55	67.73 44	85.81 64	92.35 14	75.78 19	78.33 31	86.58 80	64.01 28	94.35 36	76.05 27	87.48 40	90.79 40
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 30	81.27 30	84.50 32	89.23 40	68.76 34	90.22 24	91.94 21	75.37 20	76.64 49	91.51 24	54.29 72	94.91 25	78.44 19	83.78 56	89.83 57

API-MVS		81.99 31	81.23 31	84.26 36	90.94 23	70.18 22	91.10 15	89.32 48	71.51 41	78.66 29	88.28 58	65.26 23	95.10 22	64.74 57	91.23 22	87.51 90

PAPR		81.66 33	80.89 32	83.99 42	90.27 28	64.00 72	86.76 54	91.77 24	68.84 63	77.13 44	89.50 43	67.63 15	94.88 28	67.55 47	88.52 37	93.09 20

MAR-MVS		81.84 32	80.70 33	85.27 20	91.32 20	71.53 14	89.82 27	90.92 33	69.77 54	78.50 30	86.21 83	62.36 36	94.52 35	65.36 52	92.05 17	89.77 59
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
ACMP	74.13 6	81.51 34	80.57 34	84.36 34	89.42 33	68.69 37	89.97 26	91.50 29	74.46 25	75.04 67	90.41 36	53.82 77	94.54 33	77.56 23	82.91 60	89.86 56
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
PVSNet_Blended		80.98 37	80.34 35	82.90 56	88.85 44	65.40 61	84.43 75	92.00 18	67.62 69	78.11 33	85.05 91	66.02 20	94.27 38	71.52 39	89.50 32	89.01 66

jason		81.39 35	80.29 36	84.70 29	86.63 75	69.90 24	85.95 60	86.77 73	63.24 95	81.07 17	89.47 44	61.08 47	92.15 67	78.33 22	90.07 31	92.05 28
jason: jason.
lupinMVS		81.39 35	80.27 37	84.76 28	87.35 69	70.21 20	85.55 66	86.41 76	62.85 99	81.32 15	88.61 53	61.68 40	92.24 66	78.41 21	90.26 27	91.83 30

PVSNet_BlendedMVS		80.60 43	80.02 38	82.36 60	88.85 44	65.40 61	86.16 57	92.00 18	69.34 59	78.11 33	86.09 84	66.02 20	94.27 38	71.52 39	82.06 67	87.39 92

ACMM	73.20 8	80.78 41	79.84 39	83.58 45	89.31 37	68.37 38	89.99 25	91.60 27	70.28 50	77.25 39	89.66 41	53.37 80	93.53 49	74.24 33	82.85 61	88.85 69
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
XVG-OURS-SEG-HR		80.81 40	79.76 40	83.96 44	85.60 81	68.78 33	83.54 84	90.50 36	70.66 46	76.71 47	91.66 20	60.69 50	91.26 74	76.94 26	81.58 72	91.83 30

UGNet		80.83 39	79.59 41	84.54 31	88.04 59	68.09 40	89.42 30	88.16 63	76.95 10	76.22 52	89.46 46	49.30 95	93.94 44	68.48 44	90.31 26	91.60 32
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
114514_t		80.68 42	79.51 42	84.20 37	94.09 5	67.27 49	89.64 28	91.11 31	58.75 120	74.08 72	90.72 33	58.10 55	95.04 24	69.70 43	89.42 34	90.30 48

QAPM		80.88 38	79.50 43	85.03 22	88.01 60	68.97 31	91.59 11	92.00 18	66.63 74	75.15 65	92.16 16	57.70 56	95.45 17	63.52 58	88.76 35	90.66 42

AdaColmap		80.58 44	79.42 44	84.06 41	93.09 12	68.91 32	89.36 31	88.97 58	69.27 60	75.70 55	89.69 40	57.20 59	95.77 13	63.06 61	88.41 38	87.50 91

IterMVS-LS		80.06 48	79.38 45	82.11 61	85.89 78	63.20 76	86.79 53	89.34 47	74.19 27	75.45 57	86.72 74	66.62 17	92.39 65	72.58 37	76.86 100	90.75 41
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
test_djsdf		80.30 47	79.32 46	83.27 50	83.98 89	65.37 63	90.50 20	90.38 38	68.55 65	76.19 53	88.70 51	56.44 61	93.46 50	78.98 17	80.14 83	90.97 37

XVG-OURS		80.41 45	79.23 47	83.97 43	85.64 80	69.02 29	83.03 87	90.39 37	71.09 43	77.63 37	91.49 26	54.62 71	91.35 72	75.71 29	83.47 58	91.54 33

ab-mvs		79.51 51	78.97 48	81.14 74	88.46 56	60.91 88	83.84 79	89.24 52	70.36 49	79.03 26	88.87 50	63.23 31	90.21 89	65.12 53	82.57 65	92.28 24

PCF-MVS	73.52 7	80.38 46	78.84 49	85.01 23	87.71 63	68.99 30	83.65 81	91.46 30	63.00 97	77.77 36	90.28 37	66.10 18	95.09 23	61.40 73	88.22 39	90.94 38
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
BH-untuned		79.47 52	78.60 50	82.05 62	89.19 42	65.91 57	86.07 58	88.52 61	72.18 36	75.42 58	87.69 62	61.15 46	93.54 48	60.38 80	86.83 47	86.70 104

Vis-MVSNet (Re-imp)		78.36 59	78.45 51	78.07 101	88.64 50	51.78 126	86.70 55	79.63 118	74.14 28	75.11 66	90.83 32	61.29 45	89.75 94	58.10 94	91.60 19	92.69 22

BH-RMVSNet		79.61 50	78.44 52	83.14 52	89.38 34	65.93 56	84.95 70	87.15 69	73.56 32	78.19 32	89.79 39	56.67 60	93.36 52	59.53 86	86.74 48	90.13 51

OpenMVS	72.83 10	79.77 49	78.33 53	84.09 40	85.17 82	69.91 23	90.57 19	90.97 32	66.70 73	72.17 82	91.91 18	54.70 69	93.96 42	61.81 71	90.95 24	88.41 78

TAPA-MVS	73.13 9	79.15 53	77.94 54	82.79 57	89.59 32	62.99 80	88.16 47	91.51 28	65.77 80	77.14 43	91.09 28	60.91 49	93.21 53	50.26 119	87.05 45	92.17 25
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MVSTER		79.01 55	77.88 55	82.38 59	83.07 97	64.80 68	84.08 78	88.95 59	69.01 62	78.69 28	87.17 71	54.70 69	92.43 64	74.69 31	80.57 79	89.89 55

CDS-MVSNet		79.07 54	77.70 56	83.17 51	87.60 64	68.23 39	84.40 76	86.20 78	67.49 71	76.36 51	86.54 81	61.54 42	90.79 86	61.86 70	87.33 41	90.49 44
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
TAMVS		78.89 56	77.51 57	83.03 54	87.80 62	67.79 43	84.72 73	85.05 84	67.63 68	76.75 46	87.70 61	62.25 38	90.82 85	58.53 90	87.13 44	90.49 44

test1		78.40 57	77.40 58	81.40 69	87.60 64	63.01 77	88.39 41	89.28 49	71.63 38	75.34 60	87.28 66	54.80 66	91.11 77	62.72 62	79.57 84	90.09 52

GBi-Net		78.40 57	77.40 58	81.40 69	87.60 64	63.01 77	88.39 41	89.28 49	71.63 38	75.34 60	87.28 66	54.80 66	91.11 77	62.72 62	79.57 84	90.09 52

BH-w/o		78.21 60	77.33 60	80.84 76	88.81 47	65.13 67	84.87 71	87.85 65	69.75 55	74.52 70	84.74 93	61.34 43	93.11 57	58.24 93	85.84 53	84.27 116

FMVSNet2		78.20 61	77.21 61	81.20 72	87.60 64	62.89 81	87.47 51	89.02 56	71.63 38	75.29 64	87.28 66	54.80 66	91.10 80	62.38 66	79.38 87	89.61 61

HY-MVS	69.67 12	77.95 64	77.15 62	80.36 80	87.57 68	60.21 92	83.37 86	87.78 66	66.11 77	75.37 59	87.06 73	63.27 29	90.48 88	61.38 74	82.43 66	90.40 47

FMVSNet3		77.88 65	76.85 63	80.97 75	86.84 73	62.36 82	86.52 56	88.77 60	71.13 42	75.34 60	86.66 79	54.07 75	91.10 80	62.72 62	79.57 84	89.45 62

CNLPA		78.08 62	76.79 64	81.97 64	90.40 27	71.07 15	87.59 50	84.55 85	66.03 79	72.38 81	89.64 42	57.56 57	86.04 109	59.61 84	83.35 59	88.79 70

1112_ss		77.40 69	76.43 65	80.32 81	89.11 43	60.41 91	83.65 81	87.72 67	62.13 104	73.05 74	86.72 74	62.58 35	89.97 93	62.11 68	80.80 77	90.59 43

PAPM		77.68 66	76.40 66	81.51 67	87.29 71	61.85 85	83.78 80	89.59 44	64.74 87	71.23 87	88.70 51	62.59 34	93.66 46	52.66 113	87.03 46	89.01 66

PLC	70.83 11	78.05 63	76.37 67	83.08 53	91.88 18	67.80 42	88.19 45	89.46 46	64.33 91	69.87 98	88.38 56	53.66 79	93.58 47	58.86 88	82.73 63	87.86 86
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
TR-MVS		77.44 67	76.18 68	81.20 72	88.24 57	63.24 75	84.61 74	86.40 77	67.55 70	77.81 35	86.48 82	54.10 74	93.15 56	57.75 95	82.72 64	87.20 95

FMVSNet1		77.44 67	76.12 69	81.40 69	86.81 74	63.01 77	88.39 41	89.28 49	70.49 48	74.39 71	87.28 66	49.06 97	91.11 77	60.91 77	78.52 89	90.09 52

WTY-MVS		75.65 80	75.68 70	75.57 111	86.40 76	56.82 112	77.92 112	82.40 101	65.10 86	76.18 54	87.72 60	63.13 32	80.90 123	60.31 81	81.96 68	89.00 68

XXY-MVS		75.41 82	75.56 71	74.96 113	83.59 91	57.82 107	80.59 101	83.87 90	66.54 75	74.93 68	88.31 57	63.24 30	80.09 126	62.16 67	76.85 101	86.97 100

Test_1112_low_res		76.40 73	75.44 72	79.27 89	89.28 38	58.09 104	81.69 93	87.07 70	59.53 117	72.48 80	86.67 78	61.30 44	89.33 96	60.81 79	80.15 82	90.41 46

EPNet_dtu		75.46 81	74.86 73	77.23 104	82.57 106	54.60 120	86.89 52	83.09 94	71.64 37	66.25 114	85.86 86	55.99 63	88.04 98	54.92 104	86.55 50	89.05 65

LS3D		76.95 70	74.82 74	83.37 47	90.45 26	67.36 48	89.15 34	86.94 72	61.87 106	69.52 100	90.61 35	51.71 88	94.53 34	46.38 129	86.71 49	88.21 81

cascas		76.72 72	74.64 75	82.99 55	85.78 79	65.88 58	82.33 90	89.21 53	60.85 110	72.74 76	81.02 114	47.28 99	93.75 45	67.48 48	85.02 54	89.34 63

DP-MVS		76.78 71	74.57 76	83.42 46	93.29 10	69.46 27	88.55 39	83.70 91	63.98 94	70.20 93	88.89 49	54.01 76	94.80 30	46.66 126	81.88 70	86.01 108

LTVRE_ROB	69.57 13	76.25 75	74.54 77	81.41 68	88.60 51	64.38 71	79.24 108	89.12 54	70.76 45	69.79 99	87.86 59	49.09 96	93.20 54	56.21 102	80.16 81	86.65 105
Andreas Kuhn, Heiko Hirschmüller, Daniel Scharstein, Helmut Mayer: A TV Prior for High-Quality Scalable Multi-View Stereo Reconstruction. International Journal of Computer Vision 2016
F-COLMAP		76.38 74	74.33 78	82.50 58	89.28 38	66.95 51	88.41 40	89.03 55	64.05 93	66.83 111	88.61 53	46.78 101	92.89 59	57.48 96	78.55 88	87.67 88

XVG-ACMP-BASELINE		76.11 76	74.27 79	81.62 66	83.20 94	64.67 69	83.60 83	89.75 43	69.75 55	71.85 83	87.09 72	32.78 133	92.11 68	69.99 42	80.43 80	88.09 83

ACMH+	68.96 14	76.01 77	74.01 80	82.03 63	88.60 51	65.31 64	88.86 35	87.55 68	70.25 51	67.75 107	87.47 65	41.27 117	93.19 55	58.37 91	75.94 105	87.60 89

ACMH	67.68 16	75.89 78	73.93 81	81.77 65	88.71 49	66.61 52	88.62 38	89.01 57	69.81 53	66.78 112	86.70 77	41.95 116	91.51 71	55.64 103	78.14 93	87.17 96
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
CostFormer		75.24 83	73.90 82	79.27 89	82.65 105	58.27 103	80.80 97	82.73 98	61.57 107	75.33 63	83.13 99	55.52 64	91.07 83	64.98 56	78.34 92	88.45 76

PT_06_test		75.20 84	73.77 83	79.49 88	82.69 104	60.19 93	82.34 88	86.99 71	69.71 57	72.52 79	78.31 123	56.27 62	90.07 92	62.03 69	73.11 117	88.23 79

sss		73.60 91	73.64 84	73.51 120	82.80 103	55.01 119	76.12 116	81.69 108	62.47 103	74.68 69	85.85 87	57.32 58	78.11 130	60.86 78	80.93 75	87.39 92

IB-MVS	68.01 15	75.85 79	73.36 85	83.31 48	84.76 85	66.03 54	83.38 85	85.06 83	70.21 52	69.40 101	81.05 113	45.76 103	94.66 32	65.10 54	75.49 108	89.25 64
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
PatchFormer-LS_test		74.50 85	73.05 86	78.86 95	82.95 100	59.55 98	81.65 94	82.30 103	67.44 72	71.62 86	78.15 125	52.34 83	88.92 97	65.05 55	75.90 106	88.12 82

IterMVS		74.29 86	72.94 87	78.35 99	81.53 113	63.49 73	81.58 95	82.49 99	68.06 67	69.99 97	83.69 96	51.66 89	85.54 111	65.85 49	71.64 121	86.01 108
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
OurMVSNet-221017-0		74.26 87	72.42 88	79.80 84	83.76 90	59.59 96	85.92 62	86.64 74	66.39 76	66.96 110	87.58 64	39.46 121	91.60 70	65.76 50	69.27 125	88.22 80

Test4		73.95 89	72.20 89	79.21 91	82.91 101	58.94 100	81.25 96	82.17 104	65.21 83	71.05 89	82.44 102	44.21 108	90.17 90	63.29 59	77.28 97	88.53 74

tpmrst		72.39 101	72.13 90	73.18 121	80.54 120	49.91 132	79.91 104	79.08 120	63.11 96	71.69 85	79.95 118	55.32 65	82.77 118	65.66 51	73.89 116	86.87 101

PatchFormer_test		73.70 90	71.86 91	79.21 91	82.91 101	58.94 100	82.34 88	82.17 104	65.21 83	71.05 89	78.31 123	44.21 108	90.17 90	63.29 59	77.28 97	88.53 74

EG-PatchMatch MVS		74.04 88	71.82 92	80.71 78	84.92 84	67.42 45	85.86 63	88.08 64	66.04 78	64.22 120	83.85 94	35.10 132	92.56 62	57.44 97	80.83 76	82.16 122

tpm		72.37 103	71.71 93	74.35 116	82.19 109	52.00 124	79.22 109	77.29 123	64.56 89	72.95 75	83.68 97	51.35 90	83.26 116	58.33 92	75.80 107	87.81 87

tpm2		73.26 96	71.46 94	78.63 96	83.34 93	56.71 113	80.65 100	80.40 113	56.63 126	73.55 73	82.02 107	51.80 87	91.24 75	56.35 101	78.42 91	87.95 84

RPSCF		73.23 97	71.46 94	78.54 97	82.50 107	59.85 95	82.18 91	82.84 97	58.96 118	71.15 88	89.41 47	45.48 105	84.77 113	58.82 89	71.83 120	91.02 36

PatchmatchNet		73.12 98	71.33 96	78.49 98	83.18 95	60.85 89	79.63 105	78.57 121	64.13 92	71.73 84	79.81 120	51.20 91	85.97 110	57.40 98	76.36 102	88.66 72
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
CR-MVSNet		73.37 93	71.27 97	79.67 86	81.32 114	65.19 65	75.92 117	80.30 114	59.92 114	72.73 77	81.19 111	52.50 81	86.69 104	59.84 82	77.71 94	87.11 98

SixPastTwentyTwo		73.37 93	71.26 98	79.70 85	85.08 83	57.89 106	85.57 65	83.56 92	71.03 44	65.66 116	85.88 85	42.10 114	92.57 61	59.11 87	63.34 132	88.65 73

tpmp4_e23		73.45 92	71.17 99	80.31 82	83.55 92	59.56 97	81.88 92	82.33 102	57.94 124	70.51 92	81.62 109	51.19 92	91.63 69	53.96 108	77.51 96	89.75 60

MSDG		73.36 95	70.99 100	80.49 79	84.51 87	65.80 59	80.71 99	86.13 80	65.70 81	65.46 117	83.74 95	44.60 106	90.91 84	51.13 117	76.89 99	84.74 115

PatchMatch-RL		72.38 102	70.90 101	76.80 107	88.60 51	67.38 47	79.53 106	76.17 126	62.75 101	69.36 102	82.00 108	45.51 104	84.89 112	53.62 109	80.58 78	78.12 131

PVSNet	64.34 18	72.08 104	70.87 102	75.69 110	86.21 77	56.44 115	74.37 125	80.73 111	62.06 105	70.17 94	82.23 105	42.86 112	83.31 115	54.77 105	84.45 55	87.32 94

HyFIR		71.37 106	70.50 103	73.97 119	80.52 121	59.87 94	70.92 130	82.42 100	56.28 127	65.84 115	76.50 129	53.72 78	83.21 117	61.07 76	87.21 43	78.84 129

test_0402		72.79 100	70.44 104	79.84 83	88.13 58	65.99 55	85.93 61	84.29 87	65.57 82	67.40 109	85.49 88	46.92 100	92.61 60	35.88 138	74.38 113	80.94 124

COLMAP_ROB	66.92 17	73.01 99	70.41 105	80.81 77	87.13 72	65.63 60	88.30 44	84.19 88	62.96 98	63.80 122	87.69 62	38.04 126	92.56 62	46.66 126	74.91 110	84.24 117
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
MDTV_nov1_ep13			69.97 106		83.18 95	53.48 122	77.10 114	80.18 117	60.45 111	69.33 103	80.44 117	48.89 98	86.90 103	51.60 115	78.51 90

MIMVSNet		70.69 109	69.30 107	74.88 114	84.52 86	56.35 116	75.87 119	79.42 119	64.59 88	67.76 106	82.41 103	41.10 118	81.54 122	46.64 128	81.34 73	86.75 103

Patchmtry		70.74 108	69.16 108	75.49 112	80.72 117	54.07 121	74.94 124	80.30 114	58.34 121	70.01 95	81.19 111	52.50 81	86.54 106	53.37 110	71.09 122	85.87 110

RPMNet		71.62 105	68.94 109	79.67 86	81.32 114	65.19 65	75.92 117	78.30 122	57.60 125	72.73 77	76.45 130	52.30 84	86.69 104	48.14 123	77.71 94	87.11 98

PMMVS		69.34 115	68.67 110	71.35 125	75.67 130	62.03 84	75.17 120	73.46 132	50.00 135	68.68 104	79.05 121	52.07 85	78.13 129	61.16 75	82.77 62	73.90 134

K. test v3		71.19 107	68.51 111	79.21 91	83.04 98	57.78 108	84.35 77	76.91 125	72.90 34	62.99 124	82.86 100	39.27 122	91.09 82	61.65 72	52.66 140	88.75 71

USDC		70.33 112	68.37 112	76.21 109	80.60 119	56.23 117	79.19 110	86.49 75	60.89 109	61.29 125	85.47 89	31.78 134	89.47 95	53.37 110	76.21 103	82.94 119

tpm cat1		70.57 110	68.31 113	77.35 103	82.41 108	57.95 105	78.08 111	80.22 116	52.04 132	68.54 105	77.66 126	52.00 86	87.84 100	51.77 114	72.07 119	86.25 107

OpenMVS_ROB	64.09 19	70.56 111	68.19 114	77.65 102	80.26 122	59.41 99	85.01 69	82.96 96	58.76 119	65.43 118	82.33 104	37.63 128	91.23 76	45.34 131	76.03 104	82.32 120

CMPMVS	51.72 21	70.19 113	68.16 115	76.28 108	73.15 135	57.55 109	79.47 107	83.92 89	48.02 137	56.48 135	84.81 92	43.13 110	86.42 108	62.67 65	81.81 71	84.89 114
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
EPMVS		69.02 116	68.16 115	71.59 123	79.61 125	49.80 133	77.40 113	66.93 139	62.82 100	70.01 95	79.05 121	45.79 102	77.86 131	56.58 100	75.26 109	87.13 97

FMVSNet5		69.50 114	67.96 117	74.15 117	82.97 99	55.35 118	80.01 103	82.12 106	62.56 102	63.02 123	81.53 110	36.92 129	81.92 120	48.42 122	74.06 114	85.17 113

PatchT		68.46 118	67.85 118	70.29 127	80.70 118	43.93 137	72.47 127	74.88 130	60.15 113	70.55 91	76.57 128	49.94 93	81.59 121	50.58 118	74.83 111	85.34 112

MIMVSNet1		68.58 117	66.78 119	73.98 118	80.07 123	51.82 125	80.77 98	84.37 86	64.40 90	59.75 128	82.16 106	36.47 130	83.63 114	42.73 132	70.33 123	86.48 106

UnsupCasMVSNet_eth		67.33 120	65.99 120	71.37 124	73.48 133	51.47 128	75.16 121	85.19 82	65.20 85	60.78 127	80.93 116	42.35 113	77.20 133	57.12 99	53.69 139	85.44 111

dp		66.80 122	65.43 121	70.90 126	79.74 124	48.82 135	75.12 123	74.77 131	59.61 116	64.08 121	77.23 127	42.89 111	80.72 124	48.86 121	66.58 128	83.16 118

TinyColmap		67.30 121	64.81 122	74.76 115	81.92 110	56.68 114	80.29 102	81.49 109	60.33 112	56.27 136	83.22 98	24.77 137	87.66 101	45.52 130	69.47 124	79.95 127

TDRefinement		67.49 119	64.34 123	76.92 105	73.47 134	61.07 87	84.86 72	82.98 95	59.77 115	58.30 131	85.13 90	26.06 136	87.89 99	47.92 124	60.59 134	81.81 123

JIA-IIPM		66.32 123	62.82 124	76.82 106	77.09 128	61.72 86	65.34 136	75.38 128	58.04 123	64.51 119	62.32 139	42.05 115	86.51 107	51.45 116	69.22 126	82.21 121

LF4IMVS		64.02 124	62.19 125	69.50 129	70.90 137	53.29 123	76.13 115	77.18 124	52.65 131	58.59 129	80.98 115	23.55 138	76.52 134	53.06 112	66.66 127	78.68 130

UnsupCasMVSNet_bld		63.70 125	61.53 126	70.21 128	73.69 132	51.39 129	72.82 126	81.89 107	55.63 129	57.81 132	71.80 133	38.67 124	78.61 128	49.26 120	52.21 141	80.63 125

PVSNet_0	57.27 20	61.67 126	59.27 127	68.85 130	79.61 125	57.44 110	68.01 134	73.44 133	55.93 128	58.54 130	70.41 134	44.58 107	77.55 132	47.01 125	35.91 143	71.55 135

LP		61.36 127	57.78 128	72.09 122	75.54 131	58.53 102	67.16 135	75.22 129	51.90 133	54.13 137	69.97 135	37.73 127	80.45 125	32.74 141	55.63 137	77.29 132

MVS-HIRNet		59.14 128	57.67 129	63.57 133	81.65 112	43.50 138	71.73 128	65.06 141	39.59 140	51.43 139	57.73 140	38.34 125	82.58 119	39.53 134	73.95 115	64.62 138

testpf		56.51 130	57.58 130	53.30 140	71.99 136	41.19 140	46.89 143	69.32 137	58.06 122	52.87 138	69.45 136	27.99 135	72.73 138	59.59 85	62.07 133	45.98 144

DSMNet-mixed		57.77 129	56.90 131	60.38 134	67.70 140	35.61 143	69.18 133	53.97 144	32.30 145	57.49 133	79.88 119	40.39 120	68.57 142	38.78 135	72.37 118	76.97 133

N_pmnet		52.79 132	53.26 132	51.40 142	78.99 127	7.68 149	69.52 131	3.89 149	51.63 134	57.01 134	74.98 131	40.83 119	65.96 143	37.78 136	64.67 130	80.56 126

FPMVS		53.68 131	51.64 133	59.81 135	65.08 141	51.03 130	69.48 132	69.58 135	41.46 138	40.67 142	72.32 132	16.46 144	70.00 141	24.24 144	65.42 129	58.40 140

new_pmnet		50.91 135	50.29 134	52.78 141	68.58 139	34.94 146	63.71 138	56.63 143	39.73 139	44.95 140	65.47 137	21.93 139	58.48 145	34.98 139	56.62 136	64.92 137

HyFIR lowres test		51.79 133	50.01 135	57.11 137	68.82 138	49.21 134	60.50 139	53.26 145	34.52 141	43.77 141	64.94 138	20.34 141	71.75 139	39.87 133	64.06 131	50.39 141

ANet_high		50.57 136	46.10 136	63.99 132	48.67 145	39.13 141	70.99 129	80.85 110	61.39 108	31.18 144	57.70 141	17.02 143	73.65 137	31.22 142	15.89 149	79.18 128

no-one		51.08 134	45.79 137	66.95 131	57.92 144	50.49 131	59.63 141	76.04 127	48.04 136	31.85 143	56.10 143	19.12 142	80.08 127	36.89 137	26.52 144	70.29 136

Gipuma		45.18 137	41.86 138	55.16 139	77.03 129	51.52 127	32.50 145	80.52 112	32.46 143	27.12 145	35.02 147	9.52 146	75.50 136	22.31 145	60.21 135	38.45 146
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PMVS	37.38 22	44.16 138	40.28 139	55.82 138	40.82 148	42.54 139	65.12 137	63.99 142	34.43 142	24.48 146	57.12 142	3.92 147	76.17 135	17.10 147	55.52 138	48.75 142
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
PNet_i23d		38.26 140	35.42 140	46.79 143	58.74 142	35.48 144	59.65 140	51.25 146	32.45 144	23.44 148	47.53 145	2.04 149	58.96 144	25.60 143	18.09 148	45.92 145

wuykxyi23d		39.76 139	33.18 141	59.51 136	46.98 146	44.01 136	57.70 142	67.74 138	24.13 146	13.98 150	34.33 148	1.27 150	71.33 140	34.23 140	18.23 147	63.18 139

MVE	26.22 23	30.37 141	25.89 142	43.81 144	44.55 147	35.46 145	28.87 146	39.07 147	18.20 147	18.58 149	40.18 146	2.68 148	47.37 146	17.07 148	23.78 146	48.60 143
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
wuyk23d		16.82 142	15.94 143	19.46 146	58.74 142	31.45 147	39.22 144	3.74 150	6.84 149	6.04 151	2.70 149	1.27 150	24.29 148	10.54 149	14.40 150	2.63 148

ab-mvs-re		7.23 143	9.64 144	0.00 147	0.00 150	0.00 150	0.00 147	0.00 151	0.00 150	0.00 152	86.72 74	0.00 152	0.00 149	0.00 150	0.00 151	0.00 149

DeepMVS_CX				27.40 145	40.17 149	26.90 148		24.59 148	17.44 148	23.95 147	48.61 144	9.77 145	26.48 147	18.06 146	24.47 145	28.83 147

HQP4-MVS										77.24 40			95.11 20			91.03 35

lessismore_v0				78.97 94	81.01 116	57.15 111		65.99 140		61.16 126	82.82 101	39.12 123	91.34 73	59.67 83	46.92 142	88.43 77

LGP-MVS_train				84.50 32	89.23 40	68.76 34		91.94 21	75.37 20	76.64 49	91.51 24	54.29 72	94.91 25	78.44 19	83.78 56	89.83 57

door								69.44 136

HQP5-MVS						66.98 50

HQP-NCC					89.33 35		89.17 32		76.41 14	77.23 41

ACMP_Plane					89.33 35		89.17 32		76.41 14	77.23 41

BP-MVS														77.47 24

HQP3-MVS								92.19 16							85.99 52

ITE_SJBPF				78.22 100	81.77 111	60.57 90		83.30 93	69.25 61	67.54 108	87.20 70	36.33 131	87.28 102	54.34 106	74.62 112	86.80 102

HQP2-MVS												60.17 53

NP-MVS											90.24 38

MDTV_nov1_ep13_2view						37.79 142	75.16 121		55.10 130	66.53 113		49.34 94		53.98 107		87.94 85

ACMMP++_ref															81.95 69

ACMMP++															81.25 74

HQP-MVS												21.25 140

Test By Simon												64.33 25

door-mid								69.98 134

High-resolution multi-view benchmark