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		79.48 1	79.31 1	79.98 1	83.01 17	62.18 3	87.60 1	85.83 4	66.69 3	78.03 3	90.98 1	54.26 7	90.06 1	78.42 1	89.02 3	87.69 8
Skip Steuart: Steuart Systems R&D Blog.
DeepPCF-MVS	69.58 1	79.03 2	79.00 2	79.13 5	84.92 12	60.32 11	83.03 15	85.33 6	62.86 9	80.17 1	90.03 5	61.76 1	88.95 6	74.21 4	88.67 4	88.12 3

NCCC		78.58 3	78.31 3	79.39 4	87.51 1	62.61 2	85.20 6	84.42 11	66.73 2	74.67 5	89.38 9	55.30 6	89.18 4	74.19 5	87.34 7	86.38 12

DeepC-MVS	69.38 2	78.56 4	78.14 4	79.83 2	83.60 16	61.62 5	84.17 9	86.85 1	63.23 7	73.84 6	90.25 4	57.68 4	89.96 2	74.62 3	89.03 2	87.89 5
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
APD-MVS		78.02 5	78.04 5	77.98 10	86.44 3	60.81 8	85.52 4	84.36 12	60.61 23	79.05 2	90.30 3	55.54 5	88.32 8	73.48 7	87.03 8	84.83 25

HPM-MVS		77.28 6	76.85 6	78.54 7	85.00 9	60.81 8	82.91 17	85.08 7	62.57 10	73.09 9	89.97 6	50.90 18	87.48 12	75.30 2	86.85 9	87.33 11

DeepC-MVS_fast	68.24 3	77.25 7	76.63 9	79.12 6	86.15 5	60.86 7	84.71 7	84.85 8	61.98 15	73.06 10	88.88 11	53.72 9	89.06 5	68.27 11	88.04 5	87.42 10
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
CP-MVS		77.12 8	76.68 8	78.43 8	86.05 6	63.18 1	87.55 2	83.45 16	62.44 11	72.68 11	90.50 2	48.18 28	87.34 13	73.59 6	85.71 13	84.76 26

CSCG		76.92 9	76.75 7	77.41 11	83.96 15	59.60 13	82.95 16	86.50 2	60.78 21	75.27 4	84.83 36	60.76 2	86.56 22	67.86 12	87.87 6	86.06 13

ACMMP		76.02 10	75.33 11	78.07 9	85.20 8	61.91 4	85.49 5	84.44 10	63.04 8	69.80 16	89.74 8	45.43 36	87.16 14	72.01 8	82.87 22	85.14 22
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
PHI-MVS		75.87 11	75.36 10	77.41 11	80.62 29	55.91 35	84.28 8	85.78 5	56.08 44	73.41 7	86.58 20	50.94 17	88.54 7	70.79 9	89.71 1	87.79 7

3Dnovator+	66.72 4	75.84 12	74.57 12	79.66 3	82.40 19	59.92 12	85.83 3	86.32 3	66.92 1	67.80 32	89.24 10	42.03 50	89.38 3	64.07 26	86.50 10	89.69 1

DELS-MVS		74.76 13	74.46 13	75.65 20	77.84 55	52.25 63	75.59 64	84.17 13	63.76 6	73.15 8	82.79 50	59.58 3	86.80 19	67.24 17	86.04 12	87.89 5

HPM-MVS_fast		74.30 14	73.46 15	76.80 13	84.45 14	59.04 16	83.65 12	81.05 34	60.15 27	70.43 14	89.84 7	41.09 60	85.59 32	67.61 15	82.90 21	85.77 17

MVS_111021_HR		74.02 15	73.46 15	75.69 19	83.01 17	60.63 10	77.29 53	78.40 49	61.18 20	70.58 13	85.97 25	54.18 8	84.00 42	67.52 16	82.98 20	82.45 53

MG-MVS		73.96 16	73.89 14	74.16 35	85.65 7	49.69 75	81.59 28	81.29 32	61.45 16	71.05 12	88.11 12	51.77 13	87.73 11	61.05 31	83.09 18	85.05 24

CLD-MVS		73.33 17	72.68 17	75.29 24	78.82 45	53.33 51	78.23 46	84.79 9	61.30 19	70.41 15	81.04 70	52.41 11	87.12 15	64.61 25	82.49 23	85.41 21
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
EPNet		73.09 18	72.16 18	75.90 15	75.95 75	56.28 33	83.05 14	72.39 80	66.53 4	65.27 52	87.00 18	50.40 19	85.47 35	62.48 29	86.32 11	85.94 15

CPTT-MVS		72.78 19	72.08 19	74.87 27	84.88 13	61.41 6	84.15 10	77.86 52	55.27 50	67.51 34	88.08 14	41.93 52	81.85 57	69.04 10	80.01 31	81.35 62

LPG-MVS_test		72.74 20	71.74 21	75.76 17	80.22 32	57.51 24	82.55 23	83.40 17	61.32 17	66.67 39	87.33 16	39.15 67	86.59 20	67.70 13	77.30 47	83.19 41

PAPM_NR		72.63 21	71.80 20	75.13 25	81.72 23	53.42 50	79.91 38	83.28 20	59.14 30	66.31 42	85.90 26	51.86 12	86.06 26	57.45 39	80.62 25	85.91 16

3Dnovator	64.47 5	72.49 22	71.39 23	75.79 16	77.70 57	58.99 17	80.66 34	83.15 21	62.24 12	65.46 49	86.59 19	42.38 49	85.52 33	59.59 36	84.72 14	82.85 49

ACMP	63.53 6	72.30 23	71.20 26	75.59 22	80.28 31	57.54 23	82.74 20	82.84 24	60.58 24	65.24 53	86.18 23	39.25 66	86.03 28	66.95 19	76.79 51	83.22 39
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
Vis-MVSNet		72.18 24	71.37 24	74.61 31	81.29 24	55.41 38	80.90 32	78.28 50	60.73 22	69.23 21	88.09 13	44.36 42	82.65 53	57.68 38	81.75 24	85.77 17
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
API-MVS		72.17 25	71.41 22	74.45 33	81.95 21	57.22 26	84.03 11	80.38 35	59.89 28	68.40 24	82.33 57	49.64 20	87.83 10	51.87 46	84.16 16	78.30 86

EPP-MVSNet		72.16 26	71.31 25	74.71 28	78.68 49	49.70 73	82.10 27	81.65 28	60.40 26	65.94 45	85.84 27	51.74 14	86.37 24	55.93 40	79.55 34	88.07 4

DP-MVS Recon		72.15 27	70.73 29	76.40 14	86.57 2	57.99 21	81.15 31	82.96 22	57.03 37	66.78 38	85.56 30	44.50 40	88.11 9	51.77 48	80.23 30	83.10 44

PAPR		71.72 28	70.82 28	74.41 34	81.20 25	51.17 68	79.55 42	83.33 19	55.81 46	66.93 37	84.61 39	50.95 16	86.06 26	55.79 43	79.20 38	86.00 14

IS-MVSNet		71.57 29	71.00 27	73.27 44	78.86 44	45.63 96	80.22 36	78.69 44	64.14 5	66.46 41	87.36 15	49.30 22	85.60 31	50.26 54	83.71 17	88.59 2

MAR-MVS		71.51 30	70.15 33	75.60 21	81.84 22	59.39 14	81.38 29	82.90 23	54.90 55	68.08 26	78.70 85	47.73 29	85.51 34	51.68 50	84.17 15	81.88 58
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
MVSFormer		71.50 31	70.38 32	74.88 26	78.76 46	57.15 28	82.79 18	78.48 45	51.26 77	69.49 18	83.22 48	43.99 44	83.24 46	66.06 20	79.37 35	84.23 28

PVSNet_Blended_VisFu		71.45 32	70.39 31	74.65 29	82.01 20	58.82 18	79.93 37	80.35 36	55.09 52	65.82 48	82.16 61	49.17 24	82.64 54	60.34 33	78.62 43	82.50 52

OMC-MVS		71.40 33	70.60 30	73.78 38	76.60 70	53.15 53	79.74 41	79.78 38	58.37 32	68.75 23	86.45 21	45.43 36	80.60 68	62.58 27	77.73 45	87.58 9

PCF-MVS	61.88 8	70.95 34	69.49 36	75.35 23	77.63 59	55.71 36	76.04 62	81.81 26	50.30 81	69.66 17	85.40 33	52.51 10	84.89 37	51.82 47	80.24 29	85.45 20
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
114514_t		70.83 35	69.56 35	74.64 30	86.21 4	54.63 42	82.34 24	81.81 26	48.22 91	63.01 68	85.83 28	40.92 61	87.10 16	57.91 37	79.79 32	82.18 55

ACMM	61.98 7	70.80 36	69.73 34	74.02 36	80.59 30	58.59 19	82.68 21	82.02 25	55.46 49	67.18 35	84.39 41	38.51 73	83.17 48	60.65 32	76.10 55	80.30 71
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
QAPM		70.05 37	68.81 37	73.78 38	76.54 72	53.43 49	83.23 13	83.48 15	52.89 67	65.90 46	86.29 22	41.55 58	86.49 23	51.01 51	78.40 44	81.42 61

AdaColmap		69.99 38	68.66 39	73.97 37	84.94 10	57.83 22	82.63 22	78.71 43	56.28 43	64.34 62	84.14 43	41.57 57	87.06 17	46.45 72	78.88 40	77.02 94

jason		69.65 39	68.39 41	73.43 42	78.27 52	56.88 30	77.12 54	73.71 75	46.53 103	69.34 20	83.22 48	43.37 47	79.18 80	64.77 23	79.20 38	84.23 28
jason: jason.
lupinMVS		69.57 40	68.28 42	73.44 41	78.76 46	57.15 28	76.57 57	73.29 77	46.19 105	69.49 18	82.18 59	43.99 44	79.23 79	64.66 24	79.37 35	83.93 31

MVS_111021_LR		69.50 41	68.78 38	71.65 53	78.38 50	59.33 15	74.82 69	70.11 88	58.08 34	67.83 31	84.68 37	41.96 51	76.34 91	65.62 22	77.54 46	79.30 82

test_djsdf		69.45 42	67.74 43	74.58 32	74.57 82	54.92 40	82.79 18	78.48 45	51.26 77	65.41 50	83.49 47	38.37 75	83.24 46	66.06 20	69.25 88	85.56 19

IterMVS-LS		69.22 43	68.48 40	71.43 55	74.44 83	49.40 77	76.23 59	77.55 56	59.60 29	65.85 47	81.59 66	51.28 15	81.58 59	59.87 35	69.90 85	83.30 38
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
OpenMVS	61.03 9	68.85 44	67.56 46	72.70 49	74.26 86	53.99 45	81.21 30	81.34 31	52.70 68	62.75 70	85.55 31	38.86 70	84.14 40	48.41 64	83.01 19	79.97 75

BH-RMVSNet		68.81 45	67.42 48	72.97 46	80.11 34	52.53 59	74.26 72	76.29 64	58.48 31	68.38 25	84.20 42	42.59 48	83.83 43	46.53 71	75.91 56	82.56 50

XVG-OURS-SEG-HR		68.81 45	67.47 47	72.82 48	74.40 84	56.87 31	70.59 90	79.04 40	54.77 56	66.99 36	86.01 24	39.57 65	78.21 83	62.54 28	73.33 66	83.37 37

UGNet		68.81 45	67.39 49	73.06 45	78.33 51	54.47 43	79.77 40	75.40 68	60.45 25	63.22 66	84.40 40	32.71 98	80.91 66	51.71 49	80.56 27	83.81 35
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		68.76 48	67.37 50	72.90 47	74.32 85	57.22 26	70.09 95	78.81 42	55.24 51	67.79 33	85.81 29	36.54 86	78.28 82	62.04 30	75.74 57	83.19 41

PVSNet_Blended		68.59 49	67.72 44	71.19 59	77.03 66	50.57 71	72.51 79	81.52 29	51.91 73	64.22 64	77.77 91	49.13 25	82.87 51	55.82 41	79.58 33	80.14 74

PVSNet_BlendedMVS		68.56 50	67.72 44	71.07 62	77.03 66	50.57 71	74.50 71	81.52 29	53.66 63	64.22 64	79.72 81	49.13 25	82.87 51	55.82 41	73.92 61	79.77 79

BH-untuned		68.27 51	67.29 51	71.21 58	79.74 36	53.22 52	76.06 61	77.46 59	57.19 36	66.10 43	81.61 64	45.37 38	83.50 45	45.42 81	76.68 53	76.91 97

PAPM		67.92 52	66.69 52	71.63 54	78.09 53	49.02 78	77.09 55	81.24 33	51.04 79	60.91 76	83.98 44	47.71 30	84.99 36	40.81 100	79.32 37	80.90 65

GBi-Net		67.21 53	66.55 53	69.19 78	77.63 59	43.33 103	77.31 50	77.83 53	56.62 38	65.04 56	82.70 51	41.85 53	80.33 69	47.18 67	72.76 67	83.92 32

test1		67.21 53	66.55 53	69.19 78	77.63 59	43.33 103	77.31 50	77.83 53	56.62 38	65.04 56	82.70 51	41.85 53	80.33 69	47.18 67	72.76 67	83.92 32

MVSTER		67.16 55	65.58 60	71.88 51	70.37 109	49.70 73	70.25 94	78.45 47	51.52 74	69.16 22	80.37 76	38.45 74	82.50 55	60.19 34	71.46 78	83.44 36

FMVSNet2		66.93 56	66.31 56	68.79 82	77.63 59	42.98 106	76.11 60	77.47 57	56.62 38	65.22 55	82.17 60	41.85 53	80.18 72	47.05 70	72.72 70	83.20 40

BH-w/o		66.85 57	65.83 58	69.90 73	79.29 38	52.46 61	74.66 70	76.65 62	54.51 59	64.85 59	78.12 86	45.59 34	82.95 50	43.26 88	75.54 58	74.27 110

CDS-MVSNet		66.80 58	65.37 61	71.10 61	78.98 43	53.13 54	73.27 76	71.07 83	52.15 71	64.72 60	80.23 79	43.56 46	77.10 89	45.48 80	78.88 40	83.05 45
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
TAMVS		66.78 59	65.27 62	71.33 57	79.16 41	53.67 46	73.84 74	69.59 93	52.32 70	65.28 51	81.72 63	44.49 41	77.40 88	42.32 93	78.66 42	82.92 46

FMVSNet1		66.70 60	65.87 57	69.19 78	77.49 63	43.33 103	77.31 50	77.83 53	56.45 41	64.60 61	82.70 51	38.08 79	80.33 69	46.08 76	72.31 73	83.92 32

ab-mvs		66.65 61	66.42 55	67.37 90	76.17 74	41.73 111	70.41 92	76.14 66	53.99 61	65.98 44	83.51 46	49.48 21	76.24 92	48.60 62	73.46 64	84.14 30

TAPA-MVS	59.36 10	66.60 62	65.20 63	70.81 63	76.63 69	48.75 79	76.52 58	80.04 37	50.64 80	65.24 53	84.93 35	39.15 67	78.54 81	36.77 115	76.88 50	85.14 22
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
TR-MVS		66.59 63	65.07 64	71.17 60	79.18 40	49.63 76	73.48 75	75.20 69	52.95 66	67.90 27	80.33 78	39.81 64	83.68 44	43.20 89	73.56 63	80.20 72

FMVSNet3		66.32 64	65.61 59	68.46 86	76.48 73	42.34 109	74.98 68	77.15 61	55.83 45	65.04 56	81.16 68	39.91 63	80.14 73	47.18 67	72.76 67	82.90 48

ACMH+	57.40 11	66.12 65	64.06 65	72.30 50	77.79 56	52.83 55	80.39 35	78.03 51	57.30 35	57.47 98	82.55 54	27.68 114	84.17 39	45.54 78	69.78 86	79.90 76

cascas		65.98 66	63.42 71	73.64 40	77.26 64	52.58 58	72.26 80	77.21 60	48.56 88	61.21 75	74.60 111	32.57 100	85.82 30	50.38 53	76.75 52	82.52 51

DP-MVS		65.68 67	63.66 69	71.75 52	84.93 11	56.87 31	80.74 33	73.16 78	53.06 65	59.09 86	82.35 56	36.79 85	85.94 29	32.82 124	69.96 84	72.45 118

CNLPA		65.43 68	64.02 66	69.68 74	78.73 48	58.07 20	77.82 49	70.71 85	51.49 75	61.57 73	83.58 45	38.23 78	70.82 102	43.90 85	70.10 82	80.16 73

IB-MVS	56.42 12	65.40 69	62.73 75	73.40 43	74.89 78	52.78 56	73.09 77	75.13 70	55.69 48	58.48 95	73.73 112	32.86 97	86.32 25	50.63 52	70.11 81	81.10 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
ACMH	55.70 15	65.20 70	63.57 70	70.07 70	78.07 54	52.01 67	79.48 43	79.69 39	55.75 47	56.59 102	80.98 71	27.12 117	80.94 65	42.90 91	71.58 77	77.25 93
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
PLC	56.13 14	65.09 71	63.21 73	70.72 65	81.04 26	54.87 41	78.57 44	77.47 57	48.51 89	55.71 105	81.89 62	33.71 93	79.71 74	41.66 98	70.37 80	77.58 90
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
LS3D		64.71 72	62.50 77	71.34 56	79.72 37	55.71 36	79.82 39	74.72 72	48.50 90	56.62 101	84.62 38	33.59 94	82.34 56	29.65 130	75.23 59	75.97 99

EG-PatchMatch MVS		64.71 72	62.87 74	70.22 68	77.68 58	53.48 48	77.99 48	78.82 41	53.37 64	56.03 104	77.41 93	24.75 127	84.04 41	46.37 73	73.42 65	73.14 114

HY-MVS	56.14 13	64.55 74	63.89 67	66.55 93	74.73 80	41.02 112	69.96 96	74.43 73	49.29 85	61.66 72	80.92 72	47.43 31	76.68 90	44.91 83	71.69 76	81.94 56

XVG-ACMP-BASELINE		64.36 75	62.23 79	70.74 64	72.35 95	52.45 62	70.80 89	78.45 47	53.84 62	59.87 80	81.10 69	16.24 134	79.32 78	55.64 44	71.76 75	80.47 68

CostFormer		64.04 76	62.51 76	68.61 85	71.88 103	45.77 95	71.30 82	70.60 86	47.55 96	64.31 63	76.61 96	41.63 56	79.62 76	49.74 56	69.00 89	80.42 69

1112_ss		64.00 77	63.36 72	65.93 97	79.28 39	42.58 108	71.35 81	72.36 81	46.41 104	60.55 77	77.89 88	46.27 33	73.28 97	46.18 74	69.97 83	81.92 57

PT_06_test		63.95 78	62.37 78	68.68 83	72.07 99	47.49 86	71.03 86	75.07 71	54.94 54	61.32 74	70.62 120	40.56 62	77.68 85	49.29 57	66.43 100	80.87 66

Vis-MVSNet (Re-imp)		63.69 79	63.88 68	63.14 103	74.75 79	31.04 138	71.16 84	63.64 112	56.32 42	59.80 82	84.99 34	44.51 39	75.46 93	39.12 106	80.62 25	82.92 46

test_0402		63.25 80	61.01 85	69.96 71	80.00 35	54.37 44	76.86 56	72.02 82	54.58 58	58.71 89	80.79 75	35.00 90	84.36 38	26.41 134	64.71 107	71.15 122

LTVRE_ROB	55.42 16	63.15 81	61.23 84	68.92 81	76.57 71	47.80 84	59.92 122	76.39 63	54.35 60	58.67 90	82.46 55	29.44 108	81.49 60	42.12 94	71.14 79	77.46 91
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		63.05 82	60.87 86	69.58 76	76.99 68	53.63 47	78.12 47	76.16 65	47.97 94	52.41 119	81.61 64	27.87 112	78.11 84	40.07 103	66.66 99	77.00 95

IterMVS		62.79 83	61.27 83	67.35 91	69.37 113	52.04 66	71.17 83	68.24 99	52.63 69	59.82 81	76.91 95	37.32 82	72.36 99	52.80 45	63.19 112	77.66 89
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
tpmp4_e23		62.71 84	60.13 90	70.45 67	73.40 89	48.39 80	72.82 78	69.49 95	44.88 110	59.91 79	74.99 107	37.79 80	81.47 61	40.22 102	67.71 96	81.48 60

Test_1112_low_res		62.32 85	61.77 81	64.00 100	79.08 42	39.53 116	68.17 100	70.17 87	43.25 119	59.03 87	79.90 80	44.08 43	71.24 101	43.79 86	68.42 91	81.25 63

Test4		62.22 86	60.35 89	67.83 88	71.98 101	46.09 93	69.76 97	69.71 89	50.09 82	58.51 93	75.72 103	30.21 104	77.63 86	49.28 58	67.91 93	79.78 77

PatchFormer-LS_test		62.20 87	60.59 87	67.04 92	72.18 98	46.82 91	70.36 93	68.62 97	51.92 72	59.19 85	70.23 123	36.86 84	75.07 94	50.23 55	65.68 104	79.23 83

tpm2		62.07 88	60.10 91	67.99 87	72.79 92	43.86 102	71.05 85	66.85 101	43.14 121	62.77 69	75.39 105	38.32 76	80.80 67	41.69 97	68.88 90	79.32 81

PatchFormer_test		61.90 89	59.93 92	67.83 88	71.98 101	46.09 93	71.03 86	69.71 89	50.09 82	58.51 93	70.62 120	30.21 104	77.63 86	49.28 58	67.91 93	79.78 77

EPNet_dtu		61.90 89	61.97 80	61.68 106	72.89 91	39.78 115	75.85 63	65.62 105	55.09 52	54.56 111	79.36 82	37.59 81	67.02 110	39.80 105	76.95 49	78.25 87

MSDG		61.81 91	59.23 93	69.55 77	72.64 94	52.63 57	70.45 91	75.81 67	51.38 76	53.70 114	76.11 98	29.52 106	81.08 64	37.70 111	65.79 103	74.93 107

SixPastTwentyTwo		61.65 92	58.80 95	70.20 69	75.80 76	47.22 88	75.59 64	69.68 92	54.61 57	54.11 113	79.26 83	27.07 118	82.96 49	43.27 87	49.79 133	80.41 70

OurMVSNet-221017-0		61.37 93	58.63 97	69.61 75	72.05 100	48.06 82	73.93 73	72.51 79	47.23 99	54.74 110	80.92 72	21.49 131	81.24 63	48.57 63	56.22 124	79.53 80

COLMAP_ROB	52.97 17	61.27 94	58.81 94	68.64 84	74.63 81	52.51 60	78.42 45	73.30 76	49.92 84	50.96 121	81.51 67	23.06 129	79.40 77	31.63 128	65.85 102	74.01 113
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
XXY-MVS		60.68 95	61.67 82	57.70 118	70.43 108	38.45 118	64.19 112	66.47 102	48.05 93	63.22 66	80.86 74	49.28 23	60.47 125	45.25 82	67.28 98	74.19 111

K. test v3		60.47 96	57.11 104	70.56 66	73.74 88	48.22 81	75.10 67	62.55 117	58.27 33	53.62 115	76.31 97	27.81 113	81.59 58	47.42 65	39.18 141	81.88 58

OpenMVS_ROB	52.78 18	60.03 97	58.14 100	65.69 98	70.47 107	44.82 97	75.33 66	70.86 84	45.04 109	56.06 103	76.00 99	26.89 120	79.65 75	35.36 120	67.29 97	72.60 116

CR-MVSNet		59.91 98	57.90 101	65.96 95	69.96 111	52.07 64	65.31 106	63.15 114	42.48 123	59.36 83	74.84 108	35.83 87	70.75 103	45.50 79	64.65 108	75.06 105

PatchmatchNet		59.84 99	58.24 98	64.65 99	73.05 90	46.70 92	69.42 99	62.18 119	47.55 96	58.88 88	71.96 116	34.49 91	69.16 107	42.99 90	63.60 111	78.07 88
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
WTY-MVS		59.75 100	60.39 88	57.85 116	72.32 96	37.83 121	61.05 121	64.18 111	45.95 107	61.91 71	79.11 84	47.01 32	60.88 124	42.50 92	69.49 87	74.83 108

tpm cat1		59.25 101	56.95 107	66.15 94	72.19 97	46.96 89	68.09 101	65.76 104	40.03 128	57.81 97	70.56 122	38.32 76	74.51 96	38.26 109	61.50 119	77.00 95

RPMNet		58.70 102	56.29 110	65.96 95	69.96 111	52.07 64	65.31 106	62.15 120	43.20 120	59.36 83	70.15 125	35.37 89	70.75 103	36.42 117	64.65 108	75.06 105

PVSNet	50.76 19	58.40 103	57.39 102	61.42 108	75.53 77	44.04 101	61.43 118	63.45 113	47.04 101	56.91 99	73.61 113	27.00 119	64.76 114	39.12 106	72.40 71	75.47 103

tpmrst		58.24 104	58.70 96	56.84 120	66.97 120	34.32 129	69.57 98	61.14 122	47.17 100	58.58 92	71.60 117	41.28 59	60.41 126	49.20 60	62.84 113	75.78 100

CMPMVS	42.80 21	57.81 105	55.97 111	63.32 101	60.98 133	47.38 87	64.66 110	69.50 94	32.06 136	46.83 129	77.80 90	29.50 107	71.36 100	48.68 61	73.75 62	71.21 121
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
HyFIR		57.80 106	57.13 103	59.83 112	69.06 115	47.96 83	56.10 129	66.29 103	38.73 130	52.74 118	68.69 127	38.61 72	67.74 109	47.24 66	76.68 53	63.80 130

MIMVSNet		57.35 107	57.07 105	58.22 115	74.21 87	37.18 122	62.46 115	60.88 123	48.88 87	55.29 109	75.99 101	31.68 101	62.04 122	31.87 127	72.35 72	75.43 104

tpm		57.34 108	58.16 99	54.86 125	71.80 104	34.77 127	67.47 102	56.04 128	48.20 92	60.10 78	76.92 94	37.17 83	53.41 138	40.76 101	65.01 106	76.40 98

Patchmtry		57.16 109	56.47 109	59.23 113	69.17 114	34.58 128	62.98 114	63.15 114	44.53 113	56.83 100	74.84 108	35.83 87	68.71 108	40.03 104	60.91 120	74.39 109

USDC		56.35 110	54.24 115	62.69 104	64.74 126	40.31 114	65.05 108	73.83 74	43.93 118	47.58 127	77.71 92	15.36 135	75.05 95	38.19 110	61.81 117	72.70 115

PatchMatch-RL		56.25 111	54.55 113	61.32 110	77.06 65	56.07 34	65.57 105	54.10 133	44.13 117	53.49 117	71.27 119	25.20 125	66.78 111	36.52 116	63.66 110	61.12 131

sss		56.17 112	56.57 108	54.96 124	66.93 121	36.32 124	57.94 126	61.69 121	41.67 125	58.64 91	75.32 106	38.72 71	56.25 136	42.04 95	66.19 101	72.31 119

FMVSNet5		55.86 113	54.93 112	58.66 114	71.05 106	36.35 123	64.18 113	62.48 118	46.76 102	50.66 122	74.73 110	25.80 123	64.04 116	33.11 123	65.57 105	75.59 102

RPSCF		55.80 114	54.22 116	60.53 111	65.13 125	42.91 107	64.30 111	57.62 126	36.84 132	58.05 96	82.28 58	28.01 111	56.24 137	37.14 113	58.61 122	82.44 54

MIMVSNet1		55.17 115	54.31 114	57.77 117	70.03 110	32.01 136	65.68 104	64.81 107	49.19 86	46.75 130	76.00 99	25.53 124	64.04 116	28.65 131	62.13 116	77.26 92

TinyColmap		54.14 116	51.72 121	61.40 109	66.84 122	41.97 110	66.52 103	68.51 98	44.81 111	42.69 135	75.77 102	11.66 139	72.94 98	31.96 126	56.77 123	69.27 124

EPMVS		53.96 117	53.69 117	54.79 126	66.12 124	31.96 137	62.34 117	49.05 136	44.42 115	55.54 106	71.33 118	30.22 103	56.70 133	41.65 99	62.54 114	75.71 101

PMMVS		53.96 117	53.26 119	56.04 121	62.60 131	50.92 69	61.17 120	56.09 127	32.81 135	53.51 116	66.84 130	34.04 92	59.93 128	44.14 84	68.18 92	57.27 135

TDRefinement		53.44 119	50.72 123	61.60 107	64.31 128	46.96 89	70.89 88	65.27 106	41.78 124	44.61 132	77.98 87	11.52 140	66.36 112	28.57 132	51.59 130	71.49 120

PatchT		53.17 120	53.44 118	52.33 131	68.29 118	25.34 144	58.21 125	54.41 131	44.46 114	54.56 111	69.05 126	33.32 95	60.94 123	36.93 114	61.76 118	70.73 123

UnsupCasMVSNet_eth		53.16 121	52.47 120	55.23 122	59.45 136	33.39 132	59.43 123	69.13 96	45.98 106	50.35 124	72.32 115	29.30 109	58.26 132	42.02 96	44.30 138	74.05 112

dp		51.89 122	51.60 122	52.77 130	68.44 117	32.45 134	62.36 116	54.57 130	44.16 116	49.31 125	67.91 128	28.87 110	56.61 134	33.89 122	54.89 126	69.24 125

JIA-IIPM		51.56 123	47.68 125	63.21 102	64.61 127	50.73 70	47.71 136	58.77 125	42.90 122	48.46 126	51.72 140	24.97 126	70.24 106	36.06 119	53.89 127	68.64 126

UnsupCasMVSNet_bld		50.07 124	48.87 124	53.66 127	60.97 134	33.67 131	57.62 127	64.56 109	39.47 129	47.38 128	64.02 133	27.47 115	59.32 129	34.69 121	43.68 139	67.98 127

LP		48.51 125	45.51 127	57.52 119	62.86 130	44.53 100	52.38 133	59.84 124	38.11 131	42.81 134	61.02 134	23.23 128	63.02 118	24.10 138	45.24 137	65.02 129

PVSNet_0	43.31 20	47.46 126	45.64 126	52.92 129	67.60 119	44.65 98	54.06 132	54.64 129	41.59 126	46.15 131	58.75 136	30.99 102	58.66 131	32.18 125	24.81 144	55.46 136

MVS-HIRNet		45.52 127	44.48 129	48.65 132	68.49 116	34.05 130	59.41 124	44.50 142	27.03 139	37.96 139	50.47 142	26.16 122	64.10 115	26.74 133	59.52 121	47.82 137

testpf		44.11 128	45.40 128	40.26 140	60.52 135	27.34 139	33.26 143	54.33 132	45.87 108	41.08 136	60.26 135	16.46 133	59.14 130	46.09 75	50.68 132	34.31 144

LF4IMVS		42.95 129	42.26 130	45.04 134	48.30 142	32.50 133	54.80 130	48.49 138	28.03 138	40.51 137	70.16 124	9.24 144	43.89 142	31.63 128	49.18 134	58.72 133

FPMVS		42.18 130	41.11 131	45.39 133	58.03 137	41.01 113	49.50 134	53.81 134	30.07 137	33.71 140	64.03 132	11.69 138	52.08 139	14.01 145	55.11 125	43.09 140

ANet_high		41.38 131	37.47 134	53.11 128	39.73 145	24.45 145	56.94 128	69.69 91	47.65 95	26.04 143	52.32 138	12.44 136	62.38 121	21.80 141	10.61 149	72.49 117

no-one		40.85 132	36.09 135	55.14 123	48.55 141	38.72 117	42.15 141	62.92 116	34.60 134	23.55 144	49.74 143	12.21 137	66.16 113	26.27 135	24.84 143	60.54 132

N_pmnet		39.35 133	40.28 132	36.54 141	63.76 129	1.62 149	49.37 135	0.76 150	34.62 133	43.61 133	66.38 131	26.25 121	42.57 144	26.02 136	51.77 129	65.44 128

DSMNet-mixed		39.30 134	38.72 133	41.03 138	51.22 139	19.66 146	45.53 138	31.35 146	15.83 145	39.80 138	67.42 129	22.19 130	45.13 141	22.43 139	52.69 128	58.31 134

HyFIR lowres test		36.24 135	34.78 136	40.64 139	50.09 140	35.94 126	43.63 140	33.13 144	20.35 144	30.75 142	51.75 139	10.74 141	51.24 140	25.80 137	46.76 136	34.75 143

PMVS	28.69 22	36.22 136	33.29 138	45.02 135	36.82 146	35.98 125	54.68 131	48.74 137	26.31 140	21.02 145	51.61 141	2.88 147	60.10 127	9.99 147	47.58 135	38.99 142
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
Gipuma		34.77 137	31.91 139	43.33 137	62.05 132	37.87 120	20.39 145	67.03 100	23.23 141	18.41 146	25.84 147	4.24 145	62.73 119	14.71 144	51.32 131	29.38 146
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
new_pmnet		34.13 138	34.29 137	33.64 142	52.63 138	18.23 147	44.43 139	33.90 143	22.81 142	30.89 141	53.18 137	10.48 142	35.72 145	20.77 142	39.51 140	46.98 138

wuykxyi23d		28.12 139	22.54 141	44.87 136	34.97 147	32.11 135	37.96 142	47.31 140	13.32 146	9.29 150	23.72 148	0.45 150	56.58 135	21.85 140	13.98 146	45.93 139

PNet_i23d		27.88 140	25.99 140	33.55 143	47.54 143	25.89 141	47.24 137	32.91 145	21.44 143	15.90 147	38.09 144	0.85 149	42.76 143	16.90 143	13.03 147	32.00 145

MVE	17.77 23	21.41 141	17.77 142	32.34 144	34.34 148	25.44 143	16.11 146	24.11 147	11.19 147	13.22 148	31.92 145	1.58 148	30.95 146	10.47 146	17.03 145	40.62 141
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
wuyk23d		13.32 142	12.52 143	15.71 145	47.54 143	26.27 140	31.06 144	1.98 149	4.93 149	5.18 151	1.94 149	0.45 150	18.54 147	6.81 149	12.83 148	2.33 148

ab-mvs-re		6.49 143	8.65 144	0.00 147	0.00 150	0.00 150	0.00 147	0.00 151	0.00 150	0.00 152	77.89 88	0.00 152	0.00 149	0.00 150	0.00 151	0.00 149

DeepMVS_CX				12.03 146	17.97 149	10.91 148		10.60 148	7.46 148	11.07 149	28.36 146	3.28 146	11.29 148	8.01 148	9.74 150	13.89 147

HQP4-MVS										67.85 28			86.93 18			84.32 27

lessismore_v0				69.91 72	71.42 105	47.80 84		50.90 135		50.39 123	75.56 104	27.43 116	81.33 62	45.91 77	34.10 142	80.59 67

LGP-MVS_train				75.76 17	80.22 32	57.51 24		83.40 17	61.32 17	66.67 39	87.33 16	39.15 67	86.59 20	67.70 13	77.30 47	83.19 41

door								47.60 139

HQP5-MVS						54.94 39

HQP-NCC					80.66 27		82.31 25		62.10 13	67.85 28

ACMP_Plane					80.66 27		82.31 25		62.10 13	67.85 28

BP-MVS														67.04 18

HQP3-MVS								83.90 14							80.35 28

ITE_SJBPF				62.09 105	66.16 123	44.55 99		64.32 110	47.36 98	55.31 108	80.34 77	19.27 132	62.68 120	36.29 118	62.39 115	79.04 84

HQP2-MVS												45.46 35

NP-MVS											85.54 32

MDTV_nov1_ep13_2view						25.89 141	61.22 119		40.10 127	51.10 120		32.97 96		38.49 108		78.61 85

MDTV_nov1_ep13			57.00 106		72.73 93	38.26 119	65.02 109	64.73 108	44.74 112	55.46 107	72.48 114	32.61 99	70.47 105	37.47 112	67.75 95

ACMMP++_ref															74.07 60

ACMMP++															72.16 74

HQP-MVS												10.18 143

Test By Simon												48.33 27

door-mid								47.19 141

High-resolution multi-view benchmark