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.

DeepC-MVS_fast	98.69 1	99.49 1	99.39 1	99.77 5	99.63 28	99.59 8	99.36 42	99.46 49	99.07 4	99.79 3	99.82 22	98.85 1	99.92 15	98.68 24	99.87 12	99.82 5
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
ACMMP		99.45 2	99.32 4	99.82 1	99.89 1	99.67 4	99.62 12	99.69 13	98.12 18	99.63 11	99.84 20	98.73 6	99.96 5	98.55 29	99.83 20	99.81 6

MVS_111021_LR		99.41 3	99.33 3	99.65 12	99.77 9	99.51 13	98.94 70	99.85 4	98.82 8	99.65 10	99.74 34	98.51 9	99.80 42	98.83 19	99.89 10	99.64 28

MVS_111021_HR		99.41 3	99.32 4	99.66 9	99.72 18	99.47 14	98.95 69	99.85 4	98.82 8	99.54 18	99.73 35	98.51 9	99.74 48	98.91 13	99.88 11	99.77 11

PVSNet_Blended_VisFu		99.36 5	99.28 7	99.61 16	99.86 3	99.07 29	99.47 27	99.93 1	97.66 36	99.71 4	99.86 14	97.73 24	99.96 5	99.47 1	99.82 21	99.79 9

CSCG		99.32 6	99.32 4	99.32 38	99.85 4	98.29 48	99.71 9	99.66 15	98.11 19	99.41 27	99.80 27	98.37 13	99.96 5	98.99 8	99.96 2	99.72 16

DeepC-MVS	98.35 2	99.30 7	99.19 8	99.64 14	99.82 5	99.23 22	99.62 12	99.55 26	98.94 6	99.63 11	99.95 1	95.82 43	99.94 9	99.37 2	99.97 1	99.73 13
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
PHI-MVS		99.30 7	99.17 9	99.70 8	99.56 37	99.52 12	99.58 16	99.80 6	97.12 53	99.62 13	99.73 35	98.58 8	99.90 21	98.61 26	99.91 5	99.68 20

LS3D		99.27 9	99.12 10	99.74 6	99.18 65	99.75 1	99.56 17	99.57 21	98.45 13	99.49 22	99.85 17	97.77 23	99.94 9	98.33 34	99.84 19	99.52 40

3Dnovator	97.25 9	99.24 10	99.05 12	99.81 3	99.12 69	99.66 5	99.84 1	99.74 8	99.09 3	98.92 56	99.90 3	95.94 40	99.98 1	98.95 11	99.92 4	99.79 9

F-COLMAP		99.19 11	99.04 14	99.64 14	99.78 7	99.27 21	99.42 36	99.54 29	97.29 48	99.41 27	99.59 50	98.42 12	99.93 12	98.19 37	99.69 30	99.73 13

3Dnovator+	97.12 13	99.18 12	98.97 21	99.82 1	99.17 66	99.68 3	99.81 4	99.51 36	99.20 2	98.72 61	99.89 5	95.68 44	99.97 3	98.86 17	99.86 14	99.81 6

sss		99.17 13	99.05 12	99.53 25	99.62 29	98.97 35	99.36 42	99.62 17	97.83 30	99.67 5	99.65 43	97.37 28	99.95 8	99.19 3	99.19 39	99.68 20

DP-MVS		99.16 14	98.95 24	99.78 4	99.77 9	99.53 10	99.41 37	99.50 39	97.03 56	99.04 51	99.88 8	97.39 27	99.92 15	98.66 25	99.90 7	99.87 1

CNLPA		99.14 15	98.99 19	99.59 18	99.58 34	99.41 16	99.16 60	99.44 54	98.45 13	99.19 42	99.49 60	98.08 17	99.89 25	97.73 49	99.75 27	99.48 44

MG-MVS		99.13 16	99.02 18	99.45 34	99.57 36	98.63 44	99.07 62	99.34 66	98.99 5	99.61 14	99.82 22	97.98 19	99.87 29	97.00 63	99.80 23	99.85 3

EPP-MVSNet		99.13 16	98.99 19	99.53 25	99.65 26	99.06 30	99.81 4	99.33 67	97.43 43	99.60 15	99.88 8	97.14 30	99.84 34	99.13 6	98.94 44	99.69 18

DP-MVS Recon		99.12 18	98.95 24	99.65 12	99.74 14	99.70 2	99.27 53	99.57 21	96.40 71	99.42 26	99.68 40	98.75 4	99.80 42	97.98 42	99.72 28	99.44 48

Vis-MVSNet		99.12 18	98.97 21	99.56 22	99.78 7	99.10 26	99.68 11	99.66 15	98.49 12	99.86 1	99.87 12	94.77 51	99.84 34	99.19 3	99.41 36	99.74 12
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
TAMVS		99.12 18	99.08 11	99.24 42	99.46 41	98.55 45	99.51 24	99.46 49	98.09 21	99.45 25	99.82 22	98.34 14	99.51 69	98.70 22	98.93 45	99.67 23

anonymous		99.09 21	99.03 16	99.25 41	99.42 44	98.73 42	99.45 28	99.46 49	98.11 19	99.46 24	99.77 31	98.01 18	99.37 71	98.70 22	98.92 48	99.66 24

OMC-MVS		99.08 22	99.04 14	99.20 44	99.67 24	98.22 50	99.28 51	99.52 33	98.07 23	99.66 8	99.81 25	97.79 22	99.78 45	97.79 47	99.81 22	99.60 31

PVSNet_Blended		99.08 22	98.97 21	99.42 35	99.76 11	98.79 40	98.78 74	99.91 2	96.74 61	99.67 5	99.49 60	97.53 25	99.88 27	98.98 9	99.85 16	99.60 31

WTY-MVS		99.06 24	98.88 26	99.61 16	99.62 29	99.16 25	99.37 40	99.56 23	98.04 25	99.53 19	99.62 48	96.84 33	99.94 9	98.85 18	98.49 57	99.72 16

IS-MVSNet		99.05 25	98.87 27	99.57 20	99.73 16	99.32 17	99.75 7	99.20 72	98.02 26	99.56 17	99.86 14	96.54 36	99.67 58	98.09 40	99.13 40	99.73 13

API-MVS		99.04 26	99.03 16	99.06 47	99.40 49	99.31 19	99.55 21	99.56 23	98.54 10	99.33 34	99.39 65	98.76 3	99.78 45	96.98 64	99.78 24	98.07 83

PAPM_NR		99.04 26	98.84 29	99.66 9	99.74 14	99.44 15	99.39 39	99.38 62	97.70 34	99.28 37	99.28 72	98.34 14	99.85 33	96.96 65	99.45 35	99.69 18

PLC	97.94 4	99.02 28	98.85 28	99.53 25	99.66 25	99.01 33	99.24 58	99.52 33	96.85 59	99.27 39	99.48 62	98.25 16	99.91 19	97.76 48	99.62 32	99.65 26
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
AdaColmap		99.01 29	98.80 31	99.66 9	99.56 37	99.54 9	99.18 59	99.70 11	98.18 17	99.35 33	99.63 46	96.32 38	99.90 21	97.48 54	99.77 25	99.55 35

1112_ss		98.98 30	98.77 34	99.59 18	99.68 23	99.02 32	99.25 56	99.48 42	97.23 50	99.13 44	99.58 52	96.93 32	99.90 21	98.87 16	98.78 53	99.84 4

MSDG		98.98 30	98.80 31	99.53 25	99.76 11	99.19 23	98.75 76	99.55 26	97.25 49	99.47 23	99.77 31	97.82 21	99.87 29	96.93 66	99.90 7	99.54 36

Test_1112_low_res		98.89 32	98.66 37	99.57 20	99.69 22	98.95 36	99.03 66	99.47 48	96.98 57	99.15 43	99.23 74	96.77 34	99.89 25	98.83 19	98.78 53	99.86 2

Vis-MVSNet (Re-imp)		98.87 33	98.72 36	99.31 39	99.71 19	98.88 37	99.80 6	99.44 54	97.91 28	99.36 32	99.78 30	95.49 45	99.43 70	97.91 43	99.11 41	99.62 30

COLMAP_ROB	97.56 6	98.86 34	98.75 35	99.17 45	99.88 2	98.53 46	99.34 46	99.59 20	97.55 39	98.70 64	99.89 5	95.83 42	99.90 21	98.10 39	99.90 7	99.08 57
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
MAR-MVS		98.86 34	98.63 39	99.54 23	99.37 52	99.66 5	99.45 28	99.54 29	96.61 66	99.01 53	99.40 64	97.09 31	99.86 31	97.68 51	99.53 34	99.10 55
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
ab-mvs		98.86 34	98.63 39	99.54 23	99.64 27	99.19 23	99.44 31	99.54 29	97.77 32	99.30 35	99.81 25	94.20 54	99.93 12	99.17 5	98.82 51	99.49 42

PVSNet_BlendedMVS		98.86 34	98.80 31	99.03 48	99.76 11	98.79 40	99.28 51	99.91 2	97.42 44	99.67 5	99.37 66	97.53 25	99.88 27	98.98 9	97.29 71	98.42 78

HY-MVS	97.30 7	98.85 38	98.64 38	99.47 32	99.42 44	99.08 28	99.62 12	99.36 64	97.39 45	99.28 37	99.68 40	96.44 37	99.92 15	98.37 32	98.22 58	99.40 51

PVSNet	96.02 17	98.85 38	98.84 29	98.89 50	99.73 16	97.28 56	98.32 85	99.60 18	97.86 29	99.50 20	99.57 55	96.75 35	99.86 31	98.56 27	99.70 29	99.54 36

PatchMatch-RL		98.84 40	98.62 41	99.52 29	99.71 19	99.28 20	99.06 63	99.77 7	97.74 33	99.50 20	99.53 58	95.41 46	99.84 34	97.17 59	99.64 31	99.44 48

DeepPCF-MVS	98.18 3	98.81 41	99.37 2	97.12 82	99.60 32	91.75 87	98.61 80	99.44 54	99.35 1	99.83 2	99.85 17	98.70 7	99.81 41	99.02 7	99.91 5	99.81 6

PMMVS		98.80 42	98.62 41	99.34 37	99.27 61	98.70 43	98.76 75	99.31 68	97.34 47	99.21 41	99.07 80	97.20 29	99.82 38	98.56 27	98.87 50	99.52 40

PAPR		98.63 43	98.34 46	99.51 31	99.40 49	99.03 31	98.80 73	99.36 64	96.33 72	99.00 54	99.12 79	98.46 11	99.84 34	95.23 77	99.37 37	99.66 24

UGNet		98.54 44	98.29 49	99.29 40	99.34 54	98.46 47	99.52 22	99.68 14	98.88 7	99.07 47	99.34 67	89.92 78	99.93 12	98.92 12	99.96 2	99.49 42

OpenMVS	96.50 16	98.47 45	98.12 50	99.52 29	99.04 77	99.53 10	99.82 2	99.72 10	94.56 76	98.08 75	99.88 8	94.73 52	99.98 1	97.47 55	99.76 26	99.06 58

IterMVS-LS		98.46 46	98.42 44	98.58 54	99.59 33	98.00 51	99.37 40	99.43 59	96.94 58	99.07 47	99.59 50	97.87 20	99.03 79	98.32 35	95.62 87	98.71 62

QAPM		98.44 47	98.01 54	99.73 7	99.16 67	99.64 7	99.82 2	99.73 9	91.92 83	98.77 60	99.90 3	95.32 47	99.97 3	97.64 52	99.86 14	99.59 33

BH-RMVSNet		98.41 48	98.08 53	99.40 36	99.41 46	98.83 39	99.30 49	98.77 82	97.70 34	98.94 55	99.65 43	92.91 61	99.74 48	96.52 68	99.55 33	99.64 28

XXY-MVS		98.38 49	98.09 52	99.24 42	99.26 62	99.32 17	99.56 17	99.55 26	97.45 42	98.71 62	99.83 21	93.23 57	99.63 64	98.88 14	96.32 80	98.76 61

ACMM	97.58 5	98.37 50	98.34 46	98.48 57	99.41 46	97.10 59	99.56 17	99.45 52	98.53 11	99.04 51	99.85 17	93.00 58	99.71 53	98.74 21	97.45 65	98.64 66
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
PatchmatchNet		98.31 51	98.36 45	98.19 67	99.16 67	95.32 76	99.27 53	98.92 77	97.37 46	99.37 31	99.58 52	94.90 49	99.70 56	97.43 58	99.21 38	99.54 36
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
RPSCF		98.22 52	98.62 41	96.99 83	99.82 5	91.58 88	99.72 8	99.44 54	96.61 66	99.66 8	99.89 5	95.92 41	99.82 38	97.46 56	99.10 42	99.57 34

CLD-MVS		98.16 53	98.10 51	98.33 63	99.29 58	96.82 62	98.75 76	99.44 54	97.83 30	99.13 44	99.55 56	92.92 60	99.67 58	98.32 35	97.69 62	98.48 76
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
ACMH	97.28 8	98.10 54	97.99 55	98.44 60	99.41 46	96.96 61	99.60 15	99.56 23	98.09 21	98.15 74	99.91 2	90.87 72	99.70 56	98.88 14	97.45 65	98.67 64
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
ACMP	97.20 11	98.06 55	97.94 56	98.45 59	99.37 52	97.01 60	99.44 31	99.49 41	97.54 40	98.45 69	99.79 28	91.95 66	99.72 51	97.91 43	97.49 64	98.62 67
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
FMVSNet3		98.03 56	97.76 61	98.84 51	99.39 51	98.98 34	99.40 38	99.38 62	96.67 64	99.07 47	99.28 72	92.93 59	98.98 82	97.10 60	96.65 77	98.56 74

LTVRE_ROB	97.16 12	98.02 57	97.90 57	98.40 61	99.23 63	96.80 63	99.70 10	99.60 18	97.12 53	98.18 73	99.70 37	91.73 67	99.72 51	98.39 31	97.45 65	98.68 63
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
ACMH+	97.24 10	97.92 58	97.78 59	98.32 64	99.46 41	96.68 65	99.56 17	99.54 29	98.41 15	97.79 78	99.87 12	90.18 76	99.66 60	98.05 41	97.18 74	98.62 67

IterMVS		97.83 59	97.77 60	98.02 69	99.58 34	96.27 68	99.02 67	99.48 42	97.22 51	98.71 62	99.70 37	92.75 62	99.13 76	97.46 56	96.00 83	98.67 64

EPMVS		97.82 60	97.65 62	98.35 62	98.88 80	95.98 70	99.49 25	94.71 92	97.57 38	99.26 40	99.48 62	92.46 65	99.71 53	97.87 45	99.08 43	99.35 52

TR-MVS		97.76 61	97.41 67	98.82 52	99.06 75	97.87 52	98.87 72	98.56 87	96.63 65	98.68 65	99.22 75	92.49 64	99.65 61	95.40 76	97.79 60	98.95 59

dp		97.75 62	97.80 58	97.59 76	99.10 72	93.71 83	99.32 47	98.88 80	96.48 70	99.08 46	99.55 56	92.67 63	99.82 38	96.52 68	98.58 56	99.24 54

TAPA-MVS	97.07 15	97.74 63	97.34 69	98.94 49	99.70 21	97.53 53	99.25 56	99.51 36	91.90 84	99.30 35	99.63 46	98.78 2	99.64 62	88.09 91	99.87 12	99.65 26
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MIMVSNet		97.73 64	97.45 65	98.57 55	99.45 43	97.50 54	99.02 67	98.98 76	96.11 73	99.41 27	99.14 77	90.28 73	98.74 86	95.74 73	98.93 45	99.47 45

FMVSNet2		97.72 65	97.36 68	98.80 53	99.51 39	98.84 38	99.45 28	99.42 60	96.49 68	98.86 57	99.29 71	90.26 74	98.98 82	96.44 70	96.56 79	98.58 73

cascas		97.69 66	97.43 66	98.48 57	98.60 84	97.30 55	98.18 88	99.39 61	92.96 82	98.41 70	98.78 86	93.77 56	99.27 74	98.16 38	98.61 55	98.86 60

GBi-Net		97.68 67	97.48 63	98.29 65	99.51 39	97.26 57	99.43 34	99.48 42	96.49 68	99.07 47	99.32 69	90.26 74	98.98 82	97.10 60	96.65 77	98.62 67

PCF-MVS	97.08 14	97.66 68	97.06 72	99.47 32	99.61 31	99.09 27	98.04 89	99.25 70	91.24 85	98.51 68	99.70 37	94.55 53	99.91 19	92.76 82	99.85 16	99.42 50
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
PAPM		97.59 69	97.09 71	99.07 46	99.06 75	98.26 49	98.30 86	99.10 74	94.88 75	98.08 75	99.34 67	96.27 39	99.64 62	89.87 88	98.92 48	99.31 53

LF4IMVS		97.52 70	97.46 64	97.70 75	98.98 79	95.55 74	99.29 50	98.82 81	98.07 23	98.66 66	99.64 45	89.97 77	99.61 66	97.01 62	96.68 76	97.94 86

USDC		97.34 71	97.20 70	97.75 74	99.07 73	95.20 77	98.51 83	99.04 75	97.99 27	98.31 72	99.86 14	89.02 80	99.55 68	95.67 74	97.36 70	98.49 75

TinyColmap		97.12 72	96.89 73	97.83 73	99.07 73	95.52 75	98.57 81	98.74 85	97.58 37	97.81 77	99.79 28	88.16 84	99.56 67	95.10 78	97.21 72	98.39 80

FMVSNet1		96.84 73	96.36 74	98.29 65	99.32 55	97.26 57	99.43 34	99.48 42	95.11 74	98.55 67	99.32 69	83.95 89	98.98 82	95.81 72	96.26 81	98.62 67

test_0402		96.64 74	96.24 75	97.85 72	98.85 81	96.43 66	99.44 31	99.26 69	93.52 81	96.98 84	99.52 59	88.52 82	99.20 75	92.58 84	97.50 63	97.93 87

UnsupCasMVSNet_eth		96.44 75	96.12 77	97.40 79	98.65 83	95.65 72	99.36 42	99.51 36	97.13 52	96.04 88	98.99 83	88.40 83	98.17 88	96.71 67	90.27 91	98.40 79

FMVSNet5		96.43 76	96.19 76	97.15 81	99.11 70	95.89 71	99.32 47	99.52 33	94.47 77	98.34 71	99.07 80	87.54 85	97.07 90	92.61 83	95.72 85	98.47 77

new_pmnet		96.38 77	96.03 78	97.41 78	98.13 86	95.16 79	99.05 65	99.20 72	93.94 78	97.39 82	98.79 85	91.61 68	99.04 78	90.43 86	95.77 84	98.05 84

IB-MVS	95.67 18	96.22 78	95.44 81	98.57 55	99.21 64	96.70 64	98.65 79	97.74 90	96.71 63	97.27 83	98.54 88	86.03 86	99.92 15	98.47 30	86.30 93	99.10 55
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
PVSNet_0	94.43 19	96.09 79	95.47 80	97.94 71	99.31 56	94.34 81	97.81 90	99.70 11	97.12 53	97.46 81	98.75 87	89.71 79	99.79 44	97.69 50	81.69 95	99.68 20

MVS-HIRNet		95.75 80	95.16 82	97.51 77	99.30 57	93.69 84	98.88 71	95.78 91	85.09 92	98.78 59	92.65 94	91.29 69	99.37 71	94.85 79	99.85 16	99.46 46

MIMVSNet1		95.51 81	95.04 83	96.92 84	97.38 87	95.60 73	99.52 22	99.50 39	93.65 80	96.97 85	99.17 76	85.28 88	96.56 91	88.36 90	95.55 88	98.60 71

TDRefinement		95.42 82	94.57 85	97.97 70	89.83 91	96.11 69	99.48 26	98.75 83	96.74 61	96.68 86	99.88 8	88.65 81	99.71 53	98.37 32	82.74 94	98.09 82

N_pmnet		94.95 83	95.83 79	92.31 89	98.47 85	79.33 92	99.12 61	92.81 96	93.87 79	97.68 80	99.13 78	93.87 55	99.01 80	91.38 85	96.19 82	98.59 72

OpenMVS_ROB	92.34 20	94.38 84	93.70 86	96.41 86	97.38 87	93.17 86	99.06 63	98.75 83	86.58 90	94.84 89	98.26 90	81.53 90	99.32 73	89.01 89	97.87 59	96.76 89

CMPMVS	69.68 23	94.13 85	94.90 84	91.84 90	97.24 90	80.01 91	98.52 82	99.48 42	89.01 88	91.99 92	99.67 42	85.67 87	99.13 76	95.44 75	97.03 75	96.39 90
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
HyFIR lowres test		93.82 86	92.65 87	97.33 80	99.03 78	93.48 85	98.72 78	97.91 89	90.29 87	97.72 79	98.28 89	62.59 93	98.99 81	90.37 87	98.93 45	99.53 39

UnsupCasMVSNet_bld		93.53 87	92.51 88	96.58 85	97.38 87	93.82 82	98.24 87	99.48 42	91.10 86	93.10 90	96.66 91	74.89 91	98.37 87	94.03 80	87.71 92	97.56 88

Gipuma		90.99 88	90.15 89	93.51 87	98.73 82	90.12 89	93.98 95	99.45 52	79.32 93	92.28 91	94.91 92	69.61 92	97.98 89	87.42 92	95.67 86	92.45 92
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
FPMVS		84.93 89	85.65 90	82.75 95	86.77 92	63.39 97	98.35 84	98.92 77	74.11 95	83.39 93	98.98 84	50.85 94	92.40 93	84.54 93	94.97 89	92.46 91

PNet_i23d		79.43 90	77.68 91	84.67 92	86.18 93	71.69 96	96.50 93	93.68 93	75.17 94	71.33 94	91.18 95	32.18 96	90.62 94	78.57 95	74.34 96	91.71 94

ANet_high		77.30 91	74.86 93	84.62 93	75.88 97	77.61 93	97.63 91	93.15 95	88.81 89	64.27 96	89.29 96	36.51 95	83.93 97	75.89 96	52.31 99	92.33 93

MVE	76.82 21	76.91 92	74.31 94	84.70 91	85.38 95	76.05 94	96.88 92	93.17 94	67.39 97	71.28 95	89.01 97	21.66 100	87.69 96	71.74 98	72.29 97	90.35 95
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
PMVS	70.75 22	75.98 93	74.97 92	79.01 96	70.98 98	55.18 98	93.37 96	98.21 88	65.08 98	61.78 98	93.83 93	21.74 99	92.53 92	78.59 94	91.12 90	89.34 96
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
wuykxyi23d		74.42 94	71.19 95	84.14 94	76.16 96	74.29 95	96.00 94	92.57 97	69.57 96	63.84 97	87.49 98	21.98 97	88.86 95	75.56 97	57.50 98	89.26 97

wuyk23d		40.18 95	41.29 96	36.84 97	86.18 93	49.12 99	79.73 97	22.81 98	27.64 99	25.46 99	28.45 99	21.98 97	48.89 98	55.80 99	23.56 100	12.51 98

ab-mvs-re		8.30 96	11.06 97	0.00 98	0.00 99	0.00 100	0.00 98	0.00 99	0.00 100	0.00 100	99.58 52	0.00 101	0.00 99	0.00 100	0.00 101	0.00 99

DeepMVS_CX				93.34 88	99.29 58	82.27 90		99.22 71	85.15 91	96.33 87	99.05 82	90.97 71	99.73 50	93.57 81	97.77 61	98.01 85

ITE_SJBPF				98.08 68	99.29 58	96.37 67		98.92 77	98.34 16	98.83 58	99.75 33	91.09 70	99.62 65	95.82 71	97.40 69	98.25 81

ACMMP++															97.43 68

ACMMP++_ref															97.19 73

MDTV_nov1_ep13			98.32 48		99.11 70	94.44 80	99.27 53	98.74 85	97.51 41	99.40 30	99.62 48	94.78 50	99.76 47	97.59 53	98.81 52

MDTV_nov1_ep13_2view						95.18 78	99.35 45		96.84 60	99.58 16		95.19 48		97.82 46		99.46 46

Test By Simon												98.75 4

High-resolution multi-view benchmark