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		88.16 1	88.22 2	87.98 2	92.00 14	72.76 4	92.99 3	93.57 4	79.38 2	87.27 2	94.47 1	71.72 4	96.63 4	87.17 1	95.44 3	94.56 4
Skip Steuart: Steuart Systems R&D Blog.
DeepPCF-MVS	80.84 1	88.10 2	88.56 1	86.73 9	92.24 13	69.03 20	89.57 21	93.39 6	77.53 4	89.79 1	94.12 4	78.98 1	96.58 5	85.66 2	95.72 2	94.58 3

CP-MVS		87.11 3	86.92 3	87.68 5	94.20 2	73.86 1	93.98 1	92.82 9	76.62 9	83.68 5	94.46 2	67.93 10	95.95 9	84.20 4	94.39 6	93.23 11

DeepC-MVS	79.81 2	87.08 4	86.88 4	87.69 4	91.16 18	72.32 9	90.31 15	93.94 2	77.12 6	82.82 7	94.23 3	72.13 3	97.09 1	84.83 3	95.37 4	93.65 8
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 5	86.62 5	87.76 3	93.52 6	72.37 8	91.26 10	93.04 8	76.62 9	84.22 4	93.36 7	71.44 5	96.76 2	80.82 9	95.33 5	94.16 6
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 6	86.17 7	87.24 7	90.88 21	70.96 13	92.27 6	94.07 1	72.45 24	85.22 3	91.90 15	69.47 7	96.42 6	83.28 5	95.94 1	94.35 5

CSCG		86.41 7	86.19 6	87.07 8	92.91 10	72.48 7	90.81 12	93.56 5	73.95 21	83.16 6	91.07 20	75.94 2	95.19 12	79.94 11	94.38 7	93.55 9

ACMMP		85.89 8	85.39 8	87.38 6	93.59 5	72.63 6	92.74 4	93.18 7	76.78 8	80.73 11	93.82 5	64.33 21	96.29 7	82.67 6	90.69 20	93.23 11

3Dnovator+	77.84 4	85.48 9	84.47 10	88.51 1	91.08 19	73.49 2	93.18 2	93.78 3	80.79 1	76.66 35	93.37 6	60.40 43	96.75 3	77.20 13	93.73 8	95.29 1

MVS_111021_HR		85.14 10	84.75 9	86.32 11	91.65 16	72.70 5	85.98 48	90.33 31	76.11 12	82.08 8	91.61 17	71.36 6	94.17 32	81.02 7	92.58 10	92.08 21

CPTT-MVS		83.73 11	83.33 12	84.92 18	93.28 8	70.86 15	92.09 7	90.38 30	68.75 47	79.57 14	92.83 9	60.60 42	93.04 47	80.92 8	91.56 15	90.86 27

Vis-MVSNet		83.46 12	82.80 15	85.43 13	90.25 26	68.74 24	90.30 16	90.13 32	76.33 11	80.87 10	92.89 8	61.00 40	94.20 31	72.45 24	90.97 18	93.35 10
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
MG-MVS		83.41 13	83.45 11	83.28 36	92.74 11	62.28 68	88.17 36	89.50 34	75.22 14	81.49 9	92.74 11	66.75 13	95.11 14	72.85 21	91.58 14	92.45 17

EPP-MVSNet		83.40 14	83.02 13	84.57 21	90.13 27	64.47 54	92.32 5	90.73 29	74.45 18	79.35 16	91.10 18	69.05 9	95.12 13	72.78 22	87.22 34	94.13 7

3Dnovator	76.31 5	83.38 15	82.31 17	86.59 10	87.94 53	72.94 3	90.64 13	92.14 13	77.21 5	75.47 40	92.83 9	58.56 44	94.72 22	73.24 19	92.71 9	92.13 20

IS-MVSNet		83.15 16	82.81 14	84.18 27	89.94 28	63.30 59	91.59 8	88.46 51	79.04 3	79.49 15	92.16 12	65.10 20	94.28 28	67.71 31	91.86 12	94.95 2

DP-MVS Recon		83.11 17	82.09 19	86.15 12	94.44 1	70.92 14	88.79 26	92.20 12	70.53 32	79.17 17	91.03 22	64.12 23	96.03 8	68.39 30	90.14 22	91.50 23

PAPM_NR		83.02 18	82.41 16	84.82 20	92.47 12	66.37 39	87.93 39	91.80 17	73.82 22	77.32 30	90.66 25	67.90 11	94.90 18	70.37 27	89.48 25	93.19 13

OMC-MVS		82.69 19	81.97 20	84.85 19	88.75 41	67.42 32	87.98 38	90.87 28	74.92 15	79.72 13	91.65 16	62.19 34	93.96 33	75.26 15	86.42 41	93.16 14

PVSNet_Blended_VisFu		82.62 20	81.83 21	84.96 17	90.80 22	69.76 18	88.74 27	91.70 19	69.39 41	78.96 19	88.46 40	65.47 18	94.87 20	74.42 17	88.57 28	90.24 37

MVS_111021_LR		82.61 21	82.11 18	84.11 28	88.82 39	71.58 10	85.15 52	86.16 64	74.69 16	80.47 12	91.04 21	62.29 32	90.55 68	80.33 10	90.08 23	90.20 38

CLD-MVS		82.31 22	81.65 22	84.29 24	88.47 47	67.73 31	85.81 51	92.35 11	75.78 13	78.33 24	86.58 63	64.01 24	94.35 27	76.05 14	87.48 32	90.79 28
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
API-MVS		81.99 23	81.23 23	84.26 25	90.94 20	70.18 16	91.10 11	89.32 37	71.51 29	78.66 22	88.28 43	65.26 19	95.10 15	64.74 40	91.23 16	87.51 66

MAR-MVS		81.84 24	80.70 25	85.27 14	91.32 17	71.53 11	89.82 19	90.92 27	69.77 39	78.50 23	86.21 66	62.36 31	94.52 26	65.36 36	92.05 11	89.77 45
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 25	80.89 24	83.99 31	90.27 25	64.00 56	86.76 43	91.77 18	68.84 46	77.13 33	89.50 33	67.63 12	94.88 19	67.55 32	88.52 29	93.09 15

ACMP	74.13 6	81.51 26	80.57 26	84.36 23	89.42 30	68.69 25	89.97 18	91.50 22	74.46 17	75.04 51	90.41 27	53.82 61	94.54 24	77.56 12	82.91 46	89.86 44
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
PVSNet_Blended		80.98 27	80.34 27	82.90 42	88.85 37	65.40 47	84.43 59	92.00 14	67.62 50	78.11 26	85.05 73	66.02 16	94.27 29	71.52 25	89.50 24	89.01 51

QAPM		80.88 28	79.50 33	85.03 15	88.01 52	68.97 22	91.59 8	92.00 14	66.63 54	75.15 49	92.16 12	57.70 46	95.45 11	63.52 41	88.76 27	90.66 30

UGNet		80.83 29	79.59 31	84.54 22	88.04 51	68.09 28	89.42 22	88.16 52	76.95 7	76.22 37	89.46 34	49.30 77	93.94 35	68.48 29	90.31 21	91.60 22

ACMM	73.20 8	80.78 30	79.84 30	83.58 32	89.31 32	68.37 26	89.99 17	91.60 20	70.28 35	77.25 31	89.66 31	53.37 63	93.53 40	74.24 18	82.85 47	88.85 54
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
DELS-MVS		80.73 31	80.11 28	82.61 44	84.88 71	63.68 57	81.16 73	91.08 25	66.55 55	78.93 20	85.10 72	69.06 8	92.86 49	73.22 20	90.99 17	91.27 24

Deformable NCC		80.68 32	79.51 32	84.20 26	94.09 3	67.27 36	89.64 20	91.11 24	58.75 96	74.08 56	90.72 24	58.10 45	95.04 17	69.70 28	89.42 26	90.30 36

PVSNet_BlendedMVS		80.60 33	80.02 29	82.36 47	88.85 37	65.40 47	86.16 46	92.00 14	69.34 42	78.11 26	86.09 67	66.02 16	94.27 29	71.52 25	82.06 53	87.39 68

AdaColmap		80.58 34	79.42 34	84.06 30	93.09 9	68.91 23	89.36 23	88.97 47	69.27 43	75.70 39	89.69 30	57.20 49	95.77 10	63.06 42	88.41 30	87.50 67

PCF-MVS	73.52 7	80.38 35	78.84 37	85.01 16	87.71 55	68.99 21	83.65 64	91.46 23	63.00 74	77.77 29	90.28 28	66.10 15	95.09 16	61.40 52	88.22 31	90.94 26
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
IterMVS-LS		80.06 36	79.38 35	82.11 48	85.89 67	63.20 61	86.79 42	89.34 36	74.19 19	75.45 41	86.72 57	66.62 14	92.39 54	72.58 23	76.86 81	90.75 29
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
OpenMVS	72.83 10	79.77 37	78.33 41	84.09 29	85.17 69	69.91 17	90.57 14	90.97 26	66.70 53	72.17 65	91.91 14	54.70 56	93.96 33	61.81 51	90.95 19	88.41 58

BH-RMVSNet		79.61 38	78.44 40	83.14 38	89.38 31	65.93 42	84.95 54	87.15 58	73.56 23	78.19 25	89.79 29	56.67 50	93.36 41	59.53 62	86.74 39	90.13 39

ab-mvs		79.51 39	78.97 36	81.14 60	88.46 48	60.91 73	83.84 62	89.24 41	70.36 34	79.03 18	88.87 37	63.23 27	90.21 70	65.12 37	82.57 51	92.28 18

BH-untuned		79.47 40	78.60 38	82.05 49	89.19 35	65.91 43	86.07 47	88.52 50	72.18 25	75.42 42	87.69 47	61.15 39	93.54 39	60.38 58	86.83 38	86.70 80

TAPA-MVS	73.13 9	79.15 41	77.94 42	82.79 43	89.59 29	62.99 65	88.16 37	91.51 21	65.77 60	77.14 32	91.09 19	60.91 41	93.21 42	50.26 93	87.05 36	92.17 19
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
CDS-MVSNet		79.07 42	77.70 44	83.17 37	87.60 56	68.23 27	84.40 60	86.20 63	67.49 52	76.36 36	86.54 64	61.54 35	90.79 67	61.86 50	87.33 33	90.49 32
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
MVSTER		79.01 43	77.88 43	82.38 46	83.07 80	64.80 53	84.08 61	88.95 48	69.01 45	78.69 21	87.17 55	54.70 56	92.43 53	74.69 16	80.57 64	89.89 43

TAMVS		78.89 44	77.51 45	83.03 40	87.80 54	67.79 30	84.72 57	85.05 68	67.63 49	76.75 34	87.70 46	62.25 33	90.82 66	58.53 65	87.13 35	90.49 32

GBi-Net		78.40 45	77.40 46	81.40 55	87.60 56	63.01 62	88.39 31	89.28 38	71.63 26	75.34 44	87.28 50	54.80 53	91.11 59	62.72 43	79.57 67	90.09 40

test1		78.40 45	77.40 46	81.40 55	87.60 56	63.01 62	88.39 31	89.28 38	71.63 26	75.34 44	87.28 50	54.80 53	91.11 59	62.72 43	79.57 67	90.09 40

Vis-MVSNet (Re-imp)		78.36 47	78.45 39	78.07 79	88.64 43	51.78 100	86.70 44	79.63 96	74.14 20	75.11 50	90.83 23	61.29 38	89.75 72	58.10 69	91.60 13	92.69 16

BH-w/o		78.21 48	77.33 48	80.84 62	88.81 40	65.13 52	84.87 55	87.85 54	69.75 40	74.52 54	84.74 75	61.34 36	93.11 46	58.24 68	85.84 42	84.27 92

FMVSNet2		78.20 49	77.21 49	81.20 58	87.60 56	62.89 66	87.47 41	89.02 45	71.63 26	75.29 48	87.28 50	54.80 53	91.10 62	62.38 47	79.38 70	89.61 47

CNLPA		78.08 50	76.79 52	81.97 51	90.40 24	71.07 12	87.59 40	84.55 69	66.03 59	72.38 64	89.64 32	57.56 47	86.04 85	59.61 61	83.35 45	88.79 55

PLC	70.83 11	78.05 51	76.37 55	83.08 39	91.88 15	67.80 29	88.19 35	89.46 35	64.33 69	69.87 77	88.38 41	53.66 62	93.58 38	58.86 63	82.73 49	87.86 62
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 52	77.15 50	80.36 66	87.57 60	60.21 77	83.37 67	87.78 55	66.11 57	75.37 43	87.06 56	63.27 25	90.48 69	61.38 53	82.43 52	90.40 35

FMVSNet3		77.88 53	76.85 51	80.97 61	86.84 63	62.36 67	86.52 45	88.77 49	71.13 30	75.34 44	86.66 62	54.07 59	91.10 62	62.72 43	79.57 67	89.45 48

PAPM		77.68 54	76.40 54	81.51 53	87.29 61	61.85 70	83.78 63	89.59 33	64.74 65	71.23 68	88.70 38	62.59 29	93.66 37	52.66 87	87.03 37	89.01 51

TR-MVS		77.44 55	76.18 56	81.20 58	88.24 49	63.24 60	84.61 58	86.40 62	67.55 51	77.81 28	86.48 65	54.10 58	93.15 45	57.75 70	82.72 50	87.20 71

FMVSNet1		77.44 55	76.12 57	81.40 55	86.81 64	63.01 62	88.39 31	89.28 38	70.49 33	74.39 55	87.28 50	49.06 79	91.11 59	60.91 55	78.52 72	90.09 40

1112_ss		77.40 57	76.43 53	80.32 67	89.11 36	60.41 76	83.65 64	87.72 56	62.13 80	73.05 58	86.72 57	62.58 30	89.97 71	62.11 49	80.80 62	90.59 31

LS3D		76.95 58	74.82 61	83.37 34	90.45 23	67.36 35	89.15 24	86.94 60	61.87 82	69.52 79	90.61 26	51.71 70	94.53 25	46.38 103	86.71 40	88.21 59

DP-MVS		76.78 59	74.57 63	83.42 33	93.29 7	69.46 19	88.55 29	83.70 75	63.98 72	70.20 72	88.89 36	54.01 60	94.80 21	46.66 100	81.88 56	86.01 84

cascas		76.72 60	74.64 62	82.99 41	85.78 68	65.88 44	82.33 68	89.21 42	60.85 86	72.74 60	81.02 93	47.28 81	93.75 36	67.48 33	85.02 43	89.34 49

Test_1112_low_res		76.40 61	75.44 60	79.27 72	89.28 33	58.09 82	81.69 71	87.07 59	59.53 93	72.48 63	86.67 61	61.30 37	89.33 74	60.81 57	80.15 66	90.41 34

F-COLMAP		76.38 62	74.33 65	82.50 45	89.28 33	66.95 37	88.41 30	89.03 44	64.05 71	66.83 89	88.61 39	46.78 83	92.89 48	57.48 71	78.55 71	87.67 64

LTVRE_ROB	69.57 13	76.25 63	74.54 64	81.41 54	88.60 44	64.38 55	79.24 85	89.12 43	70.76 31	69.79 78	87.86 44	49.09 78	93.20 43	56.21 77	80.16 65	86.65 81
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+	68.96 14	76.01 64	74.01 66	82.03 50	88.60 44	65.31 49	88.86 25	87.55 57	70.25 36	67.75 86	87.47 49	41.27 96	93.19 44	58.37 66	75.94 86	87.60 65

ACMH	67.68 16	75.89 65	73.93 67	81.77 52	88.71 42	66.61 38	88.62 28	89.01 46	69.81 38	66.78 90	86.70 60	41.95 95	91.51 56	55.64 78	78.14 76	87.17 72
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
IB-MVS	68.01 15	75.85 66	73.36 70	83.31 35	84.76 72	66.03 40	83.38 66	85.06 67	70.21 37	69.40 80	81.05 92	45.76 85	94.66 23	65.10 38	75.49 88	89.25 50
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 67	75.68 58	75.57 88	86.40 65	56.82 87	77.92 89	82.40 82	65.10 64	76.18 38	87.72 45	63.13 28	80.90 98	60.31 59	81.96 54	89.00 53

XXY-MVS		75.41 68	75.56 59	74.96 90	83.59 75	57.82 84	80.59 78	83.87 74	66.54 56	74.93 52	88.31 42	63.24 26	80.09 100	62.16 48	76.85 82	86.97 76

CostFormer		75.24 69	73.90 68	79.27 72	82.65 83	58.27 81	80.80 74	82.73 80	61.57 83	75.33 47	83.13 81	55.52 51	91.07 64	64.98 39	78.34 75	88.45 57

IterMVS		74.29 70	72.94 71	78.35 77	81.53 90	63.49 58	81.58 72	82.49 81	68.06 48	69.99 76	83.69 78	51.66 71	85.54 87	65.85 34	71.64 100	86.01 84
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
EG-PatchMatch MVS		74.04 71	71.82 73	80.71 64	84.92 70	67.42 32	85.86 50	88.08 53	66.04 58	64.22 95	83.85 76	35.10 107	92.56 51	57.44 72	80.83 61	82.16 98

sss		73.60 72	73.64 69	73.51 96	82.80 82	55.01 94	76.12 93	81.69 86	62.47 79	74.68 53	85.85 68	57.32 48	78.11 104	60.86 56	80.93 60	87.39 68

tpmp4_e23		73.45 73	71.17 79	80.31 68	83.55 76	59.56 79	81.88 70	82.33 83	57.94 99	70.51 71	81.62 88	51.19 74	91.63 55	53.96 82	77.51 79	89.75 46

CR-MVSNet		73.37 74	71.27 78	79.67 70	81.32 91	65.19 50	75.92 94	80.30 92	59.92 90	72.73 61	81.19 90	52.50 64	86.69 80	59.84 60	77.71 77	87.11 74

MSDG		73.36 75	70.99 80	80.49 65	84.51 74	65.80 45	80.71 76	86.13 65	65.70 61	65.46 92	83.74 77	44.60 88	90.91 65	51.13 91	76.89 80	84.74 91

tpm2		73.26 76	71.46 75	78.63 74	83.34 77	56.71 88	80.65 77	80.40 91	56.63 101	73.55 57	82.02 86	51.80 69	91.24 57	56.35 76	78.42 74	87.95 60

RPSCF		73.23 77	71.46 75	78.54 75	82.50 84	59.85 78	82.18 69	82.84 79	58.96 94	71.15 69	89.41 35	45.48 87	84.77 89	58.82 64	71.83 99	91.02 25

PatchmatchNet		73.12 78	71.33 77	78.49 76	83.18 78	60.85 74	79.63 82	78.57 99	64.13 70	71.73 66	79.81 99	51.20 73	85.97 86	57.40 73	76.36 83	88.66 56
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
COLMAP_ROB	66.92 17	73.01 79	70.41 84	80.81 63	87.13 62	65.63 46	88.30 34	84.19 72	62.96 75	63.80 97	87.69 47	38.04 102	92.56 51	46.66 100	74.91 90	84.24 93
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
test_0402		72.79 80	70.44 83	79.84 69	88.13 50	65.99 41	85.93 49	84.29 71	65.57 62	67.40 88	85.49 69	46.92 82	92.61 50	35.88 112	74.38 93	80.94 100

tpmrst		72.39 81	72.13 72	73.18 97	80.54 96	49.91 106	79.91 81	79.08 98	63.11 73	71.69 67	79.95 97	55.32 52	82.77 93	65.66 35	73.89 96	86.87 77

PatchMatch-RL		72.38 82	70.90 81	76.80 84	88.60 44	67.38 34	79.53 83	76.17 103	62.75 77	69.36 81	82.00 87	45.51 86	84.89 88	53.62 83	80.58 63	78.12 106

tpm		72.37 83	71.71 74	74.35 93	82.19 86	52.00 98	79.22 86	77.29 101	64.56 67	72.95 59	83.68 79	51.35 72	83.26 92	58.33 67	75.80 87	87.81 63

PVSNet	64.34 18	72.08 84	70.87 82	75.69 87	86.21 66	56.44 90	74.37 102	80.73 89	62.06 81	70.17 73	82.23 84	42.86 92	83.31 91	54.77 79	84.45 44	87.32 70

RPMNet		71.62 85	68.94 88	79.67 70	81.32 91	65.19 50	75.92 94	78.30 100	57.60 100	72.73 61	76.45 105	52.30 66	86.69 80	48.14 97	77.71 77	87.11 74

Patchmtry		70.74 86	69.16 87	75.49 89	80.72 93	54.07 95	74.94 101	80.30 92	58.34 97	70.01 74	81.19 90	52.50 64	86.54 82	53.37 84	71.09 101	85.87 86

MIMVSNet		70.69 87	69.30 86	74.88 91	84.52 73	56.35 91	75.87 96	79.42 97	64.59 66	67.76 85	82.41 82	41.10 97	81.54 97	46.64 102	81.34 58	86.75 79

tpm cat1		70.57 88	68.31 91	77.35 81	82.41 85	57.95 83	78.08 88	80.22 94	52.04 106	68.54 84	77.66 102	52.00 68	87.84 76	51.77 88	72.07 98	86.25 83

OpenMVS_ROB	64.09 19	70.56 89	68.19 92	77.65 80	80.26 97	59.41 80	85.01 53	82.96 78	58.76 95	65.43 93	82.33 83	37.63 103	91.23 58	45.34 105	76.03 85	82.32 96

USDC		70.33 90	68.37 90	76.21 86	80.60 95	56.23 92	79.19 87	86.49 61	60.89 85	61.29 99	85.47 70	31.78 108	89.47 73	53.37 84	76.21 84	82.94 95

CMPMVS	51.72 21	70.19 91	68.16 93	76.28 85	73.15 109	57.55 85	79.47 84	83.92 73	48.02 110	56.48 108	84.81 74	43.13 90	86.42 84	62.67 46	81.81 57	84.89 90
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
FMVSNet5		69.50 92	67.96 95	74.15 94	82.97 81	55.35 93	80.01 80	82.12 84	62.56 78	63.02 98	81.53 89	36.92 104	81.92 95	48.42 96	74.06 94	85.17 89

PMMVS		69.34 93	68.67 89	71.35 100	75.67 105	62.03 69	75.17 97	73.46 108	50.00 108	68.68 83	79.05 100	52.07 67	78.13 103	61.16 54	82.77 48	73.90 108

EPMVS		69.02 94	68.16 93	71.59 98	79.61 100	49.80 107	77.40 90	66.93 112	62.82 76	70.01 74	79.05 100	45.79 84	77.86 105	56.58 75	75.26 89	87.13 73

MIMVSNet1		68.58 95	66.78 97	73.98 95	80.07 98	51.82 99	80.77 75	84.37 70	64.40 68	59.75 101	82.16 85	36.47 105	83.63 90	42.73 106	70.33 102	86.48 82

PatchT		68.46 96	67.85 96	70.29 102	80.70 94	43.93 111	72.47 104	74.88 106	60.15 89	70.55 70	76.57 104	49.94 75	81.59 96	50.58 92	74.83 91	85.34 88

TDRefinement		67.49 97	64.34 101	76.92 82	73.47 108	61.07 72	84.86 56	82.98 77	59.77 91	58.30 104	85.13 71	26.06 109	87.89 75	47.92 98	60.59 110	81.81 99

UnsupCasMVSNet_eth		67.33 98	65.99 98	71.37 99	73.48 107	51.47 102	75.16 98	85.19 66	65.20 63	60.78 100	80.93 95	42.35 93	77.20 107	57.12 74	53.69 114	85.44 87

TinyColmap		67.30 99	64.81 100	74.76 92	81.92 87	56.68 89	80.29 79	81.49 87	60.33 88	56.27 109	83.22 80	24.77 110	87.66 77	45.52 104	69.47 103	79.95 103

dp		66.80 100	65.43 99	70.90 101	79.74 99	48.82 109	75.12 100	74.77 107	59.61 92	64.08 96	77.23 103	42.89 91	80.72 99	48.86 95	66.58 106	83.16 94

JIA-IIPM		66.32 101	62.82 102	76.82 83	77.09 103	61.72 71	65.34 111	75.38 105	58.04 98	64.51 94	62.32 112	42.05 94	86.51 83	51.45 90	69.22 104	82.21 97

LF4IMVS		64.02 102	62.19 103	69.50 104	70.90 110	53.29 97	76.13 92	77.18 102	52.65 105	58.59 102	80.98 94	23.55 111	76.52 108	53.06 86	66.66 105	78.68 105

UnsupCasMVSNet_bld		63.70 103	61.53 104	70.21 103	73.69 106	51.39 103	72.82 103	81.89 85	55.63 103	57.81 105	71.80 108	38.67 100	78.61 102	49.26 94	52.21 115	80.63 101

PVSNet_0	57.27 20	61.67 104	59.27 105	68.85 105	79.61 100	57.44 86	68.01 110	73.44 109	55.93 102	58.54 103	70.41 109	44.58 89	77.55 106	47.01 99	35.91 116	71.55 109

MVS-HIRNet		59.14 105	57.67 106	63.57 108	81.65 89	43.50 112	71.73 105	65.06 113	39.59 113	51.43 110	57.73 113	38.34 101	82.58 94	39.53 108	73.95 95	64.62 112

DSMNet-mixed		57.77 106	56.90 107	60.38 109	67.70 113	35.61 116	69.18 109	53.97 116	32.30 118	57.49 106	79.88 98	40.39 99	68.57 115	38.78 109	72.37 97	76.97 107

FPMVS		53.68 107	51.64 109	59.81 110	65.08 114	51.03 104	69.48 108	69.58 110	41.46 111	40.67 113	72.32 107	16.46 116	70.00 114	24.24 117	65.42 107	58.40 114

N_pmnet		52.79 108	53.26 108	51.40 116	78.99 102	7.68 122	69.52 107	3.89 121	51.63 107	57.01 107	74.98 106	40.83 98	65.96 116	37.78 110	64.67 108	80.56 102

HyFIR lowres test		51.79 109	50.01 111	57.11 112	68.82 111	49.21 108	60.50 114	53.26 117	34.52 114	43.77 112	64.94 111	20.34 113	71.75 112	39.87 107	64.06 109	50.39 115

no-one		51.08 110	45.79 113	66.95 106	57.92 117	50.49 105	59.63 116	76.04 104	48.04 109	31.85 114	56.10 116	19.12 114	80.08 101	36.89 111	26.52 117	70.29 110

new_pmnet		50.91 111	50.29 110	52.78 115	68.58 112	34.94 119	63.71 113	56.63 115	39.73 112	44.95 111	65.47 110	21.93 112	58.48 118	34.98 113	56.62 112	64.92 111

ANet_high		50.57 112	46.10 112	63.99 107	48.67 118	39.13 114	70.99 106	80.85 88	61.39 84	31.18 115	57.70 114	17.02 115	73.65 111	31.22 115	15.89 122	79.18 104

Gipuma		45.18 113	41.86 114	55.16 114	77.03 104	51.52 101	32.50 119	80.52 90	32.46 116	27.12 116	35.02 120	9.52 118	75.50 110	22.31 118	60.21 111	38.45 119
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PMVS	37.38 22	44.16 114	40.28 115	55.82 113	40.82 121	42.54 113	65.12 112	63.99 114	34.43 115	24.48 117	57.12 115	3.92 119	76.17 109	17.10 120	55.52 113	48.75 116
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
wuykxyi23d		39.76 115	33.18 117	59.51 111	46.98 119	44.01 110	57.70 117	67.74 111	24.13 119	13.98 121	34.33 121	1.27 122	71.33 113	34.23 114	18.23 120	63.18 113

PNet_i23d		38.26 116	35.42 116	46.79 117	58.74 115	35.48 117	59.65 115	51.25 118	32.45 117	23.44 119	47.53 118	2.04 121	58.96 117	25.60 116	18.09 121	45.92 118

MVE	26.22 23	30.37 117	25.89 118	43.81 118	44.55 120	35.46 118	28.87 120	39.07 119	18.20 120	18.58 120	40.18 119	2.68 120	47.37 119	17.07 121	23.78 119	48.60 117
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
wuyk23d		16.82 118	15.94 119	19.46 120	58.74 115	31.45 120	39.22 118	3.74 122	6.84 122	6.04 122	2.70 122	1.27 122	24.29 121	10.54 122	14.40 123	2.63 121

ab-mvs-re		7.23 119	9.64 120	0.00 121	0.00 123	0.00 123	0.00 121	0.00 123	0.00 123	0.00 123	86.72 57	0.00 124	0.00 122	0.00 123	0.00 124	0.00 122

DeepMVS_CX				27.40 119	40.17 122	26.90 121		24.59 120	17.44 121	23.95 118	48.61 117	9.77 117	26.48 120	18.06 119	24.47 118	28.83 120

ITE_SJBPF				78.22 78	81.77 88	60.57 75		83.30 76	69.25 44	67.54 87	87.20 54	36.33 106	87.28 78	54.34 80	74.62 92	86.80 78

Test By Simon												64.33 21

ACMMP++															81.25 59

ACMMP++_ref															81.95 55

MDTV_nov1_ep13			69.97 85		83.18 78	53.48 96	77.10 91	80.18 95	60.45 87	69.33 82	80.44 96	48.89 80	86.90 79	51.60 89	78.51 73

MDTV_nov1_ep13_2view						37.79 115	75.16 98		55.10 104	66.53 91		49.34 76		53.98 81		87.94 61

NCCC					93.72 4			92.64 10

High-resolution multi-view benchmark