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	courty.	delive.	electro	facade	kicker	meadow	office	pipes	playgr.	relief	relief.	terrace	terrai.

MCST-MVS		85.75 5	86.99 9	84.31 2	94.07 1	92.80 3	88.15 3	79.10 1	85.66 18	70.72 25	76.50 28	80.45 14	82.17 2	88.35 1	87.49 2	91.63 2	97.65 1

ESAPD		87.78 1	90.56 1	84.53 1	92.88 2	93.82 1	88.95 1	76.05 4	92.95 3	80.32 2	93.12 2	86.87 1	80.88 4	85.54 10	84.01 18	88.09 32	97.62 2

DELS-MVS		79.49 26	79.84 35	79.08 23	88.26 32	92.49 4	84.12 20	70.63 22	65.27 71	69.60 31	61.29 53	66.50 53	72.75 33	88.07 2	88.03 1	89.13 11	97.22 3
Christian Sormann, Emanuele Santellani, Mattia Rossi, Andreas Kuhn, Friedrich Fraundorfer: DELS-MVS: Deep Epipolar Line Search for Multi-View Stereo. Winter Conference on Applications of Computer Vision (WACV), 2023
CNVR-MVS		85.96 4	87.58 7	84.06 3	92.58 4	92.40 6	87.62 4	77.77 2	88.44 10	75.93 12	79.49 21	81.97 11	81.65 3	87.04 5	86.58 3	88.79 14	97.18 4

PVSNet_BlendedMVS		76.84 45	78.47 38	74.95 46	82.37 52	89.90 25	75.45 67	65.45 54	74.99 48	70.66 26	63.07 46	58.27 82	67.60 65	84.24 18	81.70 36	88.18 28	97.10 5

PVSNet_Blended		76.84 45	78.47 38	74.95 46	82.37 52	89.90 25	75.45 67	65.45 54	74.99 48	70.66 26	63.07 46	58.27 82	67.60 65	84.24 18	81.70 36	88.18 28	97.10 5

APDe-MVS		86.37 3	88.41 4	84.00 4	91.43 9	91.83 10	88.34 2	74.67 5	91.19 4	81.76 1	91.13 3	81.94 12	80.07 5	83.38 22	82.58 29	87.69 36	96.78 7

DeepPCF-MVS	76.94 1	83.08 14	87.77 6	77.60 30	90.11 14	90.96 14	78.48 49	72.63 17	93.10 2	65.84 37	80.67 19	81.55 13	74.80 23	85.94 9	85.39 8	83.75 150	96.77 8

CHOSEN 1792x2688		72.55 60	71.98 66	73.22 54	86.57 41	92.41 5	75.63 63	66.77 44	62.08 76	52.32 75	30.27 200	50.74 110	66.14 70	86.22 7	85.41 7	91.90 1	96.75 9

SMA-MVS		84.91 7	87.95 5	81.36 10	91.75 7	90.84 15	86.35 8	73.36 13	90.22 7	72.81 18	80.70 18	85.67 2	76.69 14	86.06 8	86.14 5	87.20 49	96.05 10

abl_6				79.06 24	89.68 19	92.14 8	77.70 54	69.68 28	86.87 14	71.88 20	74.29 30	80.06 16	76.56 16			88.84 13	95.82 11

MVS_111021_HR		77.42 41	78.40 40	76.28 36	86.95 38	90.68 16	77.41 56	70.56 25	66.21 65	62.48 46	66.17 42	63.98 60	72.08 38	82.87 26	83.15 24	88.24 27	95.71 12

TSAR-MVS + GP.		82.27 18	85.98 14	77.94 28	80.72 66	88.25 38	81.12 38	67.71 40	87.10 12	73.31 15	85.23 11	83.68 5	76.64 15	80.43 51	81.47 40	88.15 30	95.66 13

HPM-MVS++		85.64 6	88.43 3	82.39 7	92.65 3	90.24 21	85.83 10	74.21 6	90.68 6	75.63 13	86.77 9	84.15 4	78.68 9	86.33 6	85.26 9	87.32 43	95.60 14

CANet		80.90 23	82.93 24	78.53 26	86.83 40	92.26 7	81.19 37	66.95 43	81.60 31	69.90 28	66.93 39	74.80 26	76.79 13	84.68 15	84.77 15	89.50 8	95.50 15

APD-MVS		84.83 8	87.00 8	82.30 8	89.61 20	89.21 29	86.51 7	73.64 11	90.98 5	77.99 7	89.89 5	80.04 17	79.18 7	82.00 38	81.37 41	86.88 53	95.49 16
Yuesong Wang, Zhaojie Zeng and etc.: Adaptive Patch Deformation for Textureless-Resilient Multi-View Stereo. CVPR2023
NCCC		84.16 11	85.46 16	82.64 6	92.34 5	90.57 18	86.57 6	76.51 3	86.85 15	72.91 17	77.20 27	78.69 20	79.09 8	84.64 16	84.88 14	88.44 22	95.41 17

CSCG		82.90 15	84.52 18	81.02 13	91.85 6	93.43 2	87.14 5	74.01 9	81.96 28	76.14 10	70.84 32	82.49 8	69.71 50	82.32 34	85.18 11	87.26 45	95.40 18

train_agg		83.35 13	86.93 11	79.17 22	89.70 18	88.41 35	85.60 14	72.89 16	86.31 16	66.58 36	90.48 4	82.24 10	73.06 32	83.10 24	82.64 28	87.21 48	95.30 19

GG-mvs-BLEND		54.54 189	77.58 42	27.67 226	0.03 238	90.09 24	77.20 57	0.02 236	66.83 63	0.05 241	59.90 57	73.33 30	0.04 236	78.40 65	79.30 56	88.65 16	95.20 20

DeepC-MVS	74.46 3	80.30 25	81.05 31	79.42 19	87.42 36	88.50 34	83.23 23	73.27 14	82.78 25	71.01 24	62.86 48	69.93 46	74.80 23	84.30 17	84.20 17	86.79 55	94.77 21
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
PVSNet_Blended_VisFu		71.76 64	73.54 62	69.69 68	79.01 72	87.16 46	72.05 83	61.80 86	56.46 93	59.66 60	53.88 72	62.48 63	59.08 129	81.17 45	78.90 58	86.53 59	94.74 22

TSAR-MVS + MP.		84.39 9	86.58 12	81.83 9	88.09 34	86.47 53	85.63 12	73.62 12	90.13 8	79.24 4	89.67 6	82.99 7	77.72 11	81.22 44	80.92 49	86.68 56	94.66 23

SD-MVS		84.31 10	86.96 10	81.22 11	88.98 26	88.68 32	85.65 11	73.85 10	89.09 9	79.63 3	87.34 8	84.84 3	73.71 28	82.66 28	81.60 38	85.48 105	94.51 24

DI_MVS_plusplus_trai		73.94 54	74.85 57	72.88 55	76.57 91	86.80 48	80.41 42	61.47 89	62.35 75	59.44 61	47.91 91	68.12 48	72.24 36	82.84 27	81.50 39	87.15 50	94.42 25

SteuartSystems-ACMMP		82.51 16	85.35 17	79.20 21	90.25 12	89.39 28	84.79 16	70.95 20	82.86 24	68.32 33	86.44 10	77.19 21	73.07 31	83.63 21	83.64 21	87.82 33	94.34 26
Skip Steuart: Steuart Systems R&D Blog.
MVS_0304		79.43 28	82.20 26	76.20 37	84.22 47	91.79 11	81.82 32	63.81 64	76.83 45	61.71 48	66.37 41	75.52 25	76.38 17	85.54 10	85.03 12	89.28 10	94.32 27

PHI-MVS		79.43 28	84.06 20	74.04 50	86.15 42	91.57 13	80.85 41	68.90 35	82.22 27	51.81 78	78.10 23	74.28 27	70.39 47	84.01 20	84.00 19	86.14 66	94.24 28

ACMMP_Plus		83.54 12	86.37 13	80.25 16	89.57 21	90.10 23	85.27 15	71.66 18	87.38 11	73.08 16	84.23 13	80.16 15	75.31 19	84.85 14	83.64 21	86.57 57	94.21 29

MP-MVS		80.94 22	83.49 21	77.96 27	88.48 27	88.16 39	82.82 27	69.34 31	80.79 34	69.67 29	82.35 15	77.13 22	71.60 41	80.97 48	80.96 48	85.87 86	94.06 30

PCF-MVS	70.85 4	75.73 48	76.55 52	74.78 49	83.67 48	88.04 42	81.47 33	70.62 24	69.24 61	57.52 65	60.59 56	69.18 47	70.65 45	77.11 72	77.65 70	84.75 131	94.01 31
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
DeepC-MVS_fast	75.41 2	81.69 19	82.10 28	81.20 12	91.04 11	87.81 43	83.42 22	74.04 8	83.77 22	71.09 23	66.88 40	72.44 32	79.48 6	85.08 12	84.97 13	88.12 31	93.78 32
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
TSAR-MVS + ACMM		81.59 21	85.84 15	76.63 34	89.82 17	86.53 52	86.32 9	66.72 45	85.96 17	65.43 38	88.98 7	82.29 9	67.57 67	82.06 37	81.33 42	83.93 148	93.75 33

HyFIR lowres test		68.39 78	68.28 88	68.52 76	80.85 63	88.11 40	71.08 110	58.09 114	54.87 104	47.80 92	27.55 205	55.80 91	64.97 74	79.11 58	79.14 57	88.31 25	93.35 34

test-LLR		68.23 79	71.61 70	64.28 114	71.37 123	81.32 94	63.98 157	61.03 92	58.62 84	42.96 124	52.74 73	61.65 69	57.74 135	75.64 90	78.09 68	88.61 18	93.21 35

TESTMET0.1,1		67.38 85	71.61 70	62.45 138	66.05 171	81.32 94	63.98 157	55.36 149	58.62 84	42.96 124	52.74 73	61.65 69	57.74 135	75.64 90	78.09 68	88.61 18	93.21 35

zzz-MVS		81.65 20	83.10 22	79.97 18	88.14 33	87.62 44	83.96 21	69.90 26	86.92 13	77.67 9	72.47 31	78.74 19	74.13 27	81.59 42	81.15 45	86.01 73	93.19 37

HQP-MVS		78.26 35	80.91 32	75.17 45	85.67 44	84.33 68	83.01 25	69.38 30	79.88 37	55.83 68	79.85 20	64.90 58	70.81 44	82.46 30	81.78 35	86.30 62	93.18 38

IB-MVS	64.48 11	69.02 75	68.97 83	69.09 73	81.75 57	89.01 30	64.50 151	64.91 57	56.65 91	62.59 45	47.89 92	45.23 122	51.99 152	69.18 173	81.88 34	88.77 15	92.93 39
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
CDPH-MVS		79.39 31	82.13 27	76.19 38	89.22 25	88.34 36	84.20 19	71.00 19	79.67 38	56.97 67	77.77 24	72.24 36	68.50 60	81.33 43	82.74 26	87.23 46	92.84 40

CostFormer		72.18 61	73.90 60	70.18 67	79.47 69	86.19 58	76.94 59	48.62 190	66.07 67	60.40 58	54.14 70	65.82 54	67.98 62	75.84 88	76.41 77	87.67 37	92.83 41

QAPM		77.50 40	77.43 43	77.59 31	91.52 8	92.00 9	81.41 35	70.63 22	66.22 64	58.05 64	54.70 66	71.79 38	74.49 26	82.46 30	82.04 31	89.46 9	92.79 42

CP-MVS		79.44 27	81.51 30	77.02 33	86.95 38	85.96 59	82.00 29	68.44 37	81.82 29	67.39 34	77.43 25	73.68 28	71.62 40	79.56 56	79.58 54	85.73 95	92.51 43

HFP-MVS		82.48 17	84.12 19	80.56 14	90.15 13	87.55 45	84.28 18	69.67 29	85.22 19	77.95 8	84.69 12	75.94 24	75.04 21	81.85 39	81.17 44	86.30 62	92.40 44

canonicalmvs		77.65 38	79.59 36	75.39 42	81.52 58	89.83 27	81.32 36	60.74 99	80.05 36	66.72 35	68.43 36	65.09 56	74.72 25	78.87 60	82.73 27	87.32 43	92.16 45

HSP-MVS		86.82 2	89.95 2	83.16 5	89.38 22	91.60 12	85.63 12	74.15 7	94.20 1	75.52 14	94.99 1	83.21 6	85.96 1	87.67 3	85.88 6	88.32 24	92.13 46

EPNet		79.28 32	82.25 25	75.83 40	88.31 31	90.14 22	79.43 47	68.07 38	81.76 30	61.26 50	77.26 26	70.08 45	70.06 48	82.43 32	82.00 33	87.82 33	92.09 47
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
diffmvs		73.50 55	75.66 55	70.97 63	74.96 103	86.71 49	77.16 58	57.42 129	71.12 54	60.43 57	57.20 61	70.40 44	68.79 59	76.11 85	76.05 81	87.10 51	92.06 48

CPTT-MVS		75.43 49	77.13 47	73.44 52	81.43 59	82.55 77	80.96 40	64.35 59	77.95 43	61.39 49	69.20 35	70.94 41	69.38 55	73.89 109	73.32 135	83.14 161	92.06 48

test-mter		64.06 113	69.24 81	58.01 165	59.07 199	77.40 142	59.13 183	48.11 193	55.64 99	39.18 148	51.56 77	58.54 78	55.38 143	73.52 114	76.00 82	87.22 47	92.05 50

MVS_111021_LR		74.26 52	75.95 53	72.27 58	79.43 70	85.04 63	72.71 79	65.27 56	70.92 55	63.58 42	69.32 34	60.31 74	69.43 53	77.01 73	77.15 71	83.22 157	91.93 51

gg-mvs-nofinetune		62.34 135	66.19 100	57.86 169	76.15 94	88.61 33	71.18 107	41.24 219	25.74 220	13.16 223	22.91 215	63.97 61	54.52 147	85.06 13	85.25 10	90.92 3	91.78 52

ACMMPR		80.62 24	82.98 23	77.87 29	88.41 28	87.05 47	83.02 24	69.18 32	83.91 21	68.35 32	82.89 14	73.64 29	72.16 37	80.78 49	81.13 46	86.10 67	91.43 53

MVS_Test		75.22 50	76.69 50	73.51 51	79.30 71	88.82 31	80.06 44	58.74 109	69.77 58	57.50 66	59.78 58	61.35 71	75.31 19	82.07 36	83.60 23	90.13 5	91.41 54

X-MVS		78.16 36	80.55 33	75.38 43	87.99 35	86.27 55	81.05 39	68.98 33	78.33 40	61.07 52	75.25 29	72.27 33	67.52 68	80.03 53	80.52 53	85.66 102	91.20 55

MVSTER		76.92 44	79.92 34	73.42 53	74.98 101	82.97 73	78.15 50	63.41 67	78.02 41	64.41 41	67.54 37	72.80 31	71.05 43	83.29 23	83.73 20	88.53 21	91.12 56

UGNet		67.57 83	71.69 69	62.76 134	69.88 131	82.58 76	66.43 144	58.64 110	54.71 105	51.87 77	61.74 50	62.01 68	45.46 180	74.78 97	74.99 92	84.24 142	91.02 57
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
DWT-MVSNet_training		72.81 58	73.98 58	71.45 62	81.26 60	86.37 54	72.08 82	59.82 106	69.13 62	58.15 63	54.71 65	61.33 73	67.81 64	76.86 74	78.63 60	89.59 6	90.86 58

ACMMP		77.61 39	79.59 36	75.30 44	85.87 43	85.58 60	81.42 34	67.38 42	79.38 39	62.61 44	78.53 22	65.79 55	68.80 58	78.56 63	78.50 63	85.75 91	90.80 59
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
3Dnovator+	70.16 6	77.87 37	77.29 45	78.55 25	89.25 24	88.32 37	80.09 43	67.95 39	74.89 50	71.83 21	52.05 76	70.68 42	76.27 18	82.27 35	82.04 31	85.92 79	90.77 60

PGM-MVS		79.42 30	81.84 29	76.60 35	88.38 30	86.69 50	82.97 26	65.75 51	80.39 35	64.94 39	81.95 17	72.11 37	71.41 42	80.45 50	80.55 52	86.18 64	90.76 61

Effi-MVS+		70.42 67	71.23 72	69.47 69	78.04 76	85.24 62	75.57 65	58.88 108	59.56 82	48.47 89	52.73 75	54.94 96	69.69 51	78.34 66	77.06 72	86.18 64	90.73 62

tpmrst		67.15 87	68.12 90	66.03 92	76.21 93	80.98 97	71.27 104	45.05 202	60.69 80	50.63 83	46.95 105	54.15 99	65.30 72	71.80 150	71.77 153	87.72 35	90.48 63

CLD-MVS		77.36 42	77.29 45	77.45 32	82.21 54	88.11 40	81.92 30	68.96 34	77.97 42	69.62 30	62.08 49	59.44 76	73.57 29	81.75 40	81.27 43	88.41 23	90.39 64
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
EPP-MVSNet		67.58 82	71.10 73	63.48 123	75.71 97	83.35 72	66.85 140	57.83 117	53.02 109	41.15 139	55.82 63	67.89 50	56.01 141	74.40 100	72.92 144	83.33 155	90.30 65

CANet_DTU		72.84 57	76.63 51	68.43 77	76.81 89	86.62 51	75.54 66	54.71 157	72.06 52	43.54 112	67.11 38	58.46 79	72.40 35	81.13 47	80.82 51	87.57 38	90.21 66

MSLP-MVS++		78.57 33	77.33 44	80.02 17	88.39 29	84.79 65	84.62 17	66.17 49	75.96 47	78.40 5	61.59 51	71.47 39	73.54 30	78.43 64	78.88 59	88.97 12	90.18 67

3Dnovator	70.49 5	78.42 34	76.77 49	80.35 15	91.43 9	90.27 20	81.84 31	70.79 21	72.10 51	71.95 19	50.02 83	67.86 51	77.47 12	82.89 25	84.24 16	88.61 18	89.99 68

thres100view900		67.14 88	66.09 103	68.38 78	77.70 79	83.84 71	74.52 73	66.33 48	49.16 120	43.40 117	43.24 115	41.34 132	62.59 86	79.31 57	75.92 83	85.73 95	89.81 69

thres200		65.58 97	64.74 111	66.56 90	77.52 84	81.61 85	73.44 78	62.95 73	46.23 141	42.45 134	42.76 119	41.18 137	58.12 133	76.24 80	75.59 87	84.89 121	89.58 70

tpm cat1		67.47 84	67.05 96	67.98 79	76.63 90	81.51 91	74.49 75	47.65 195	61.18 78	61.12 51	42.51 124	53.02 104	64.74 77	70.11 165	71.50 156	83.22 157	89.49 71

IS_MVSNet		67.29 86	71.98 66	61.82 144	76.92 87	84.32 69	65.90 148	58.22 112	55.75 98	39.22 147	54.51 68	62.47 64	45.99 177	78.83 61	78.52 62	84.70 132	89.47 72

tpm		64.85 103	66.02 104	63.48 123	74.52 105	78.38 131	70.98 112	44.99 204	51.61 111	43.28 119	47.66 95	53.18 102	60.57 104	70.58 159	71.30 165	86.54 58	89.45 73

gm-plane-assit		54.99 184	57.99 179	51.49 194	69.27 137	54.42 219	32.32 223	42.59 211	21.18 226	13.71 221	23.61 211	43.84 126	60.21 112	87.09 4	86.55 4	90.81 4	89.28 74

OpenMVS	67.62 8	74.92 51	73.91 59	76.09 39	90.10 15	90.38 19	78.01 51	66.35 47	66.09 66	62.80 43	46.33 108	64.55 59	71.77 39	79.92 54	80.88 50	87.52 39	89.20 75

v1144		63.00 129	62.39 138	63.70 119	67.72 161	80.27 113	71.23 106	56.40 136	42.51 167	40.81 141	38.12 166	37.73 172	60.42 110	74.46 99	74.55 102	85.64 103	89.12 76

v7		63.61 118	63.02 126	64.29 113	67.88 158	80.32 112	71.60 100	56.63 135	45.37 145	42.84 129	38.54 157	38.91 168	61.05 99	74.39 101	74.52 104	85.75 91	89.10 77

ACMP	68.86 7	72.15 62	72.25 65	72.03 59	80.96 62	80.87 99	77.93 52	64.13 61	69.29 59	60.79 55	64.04 44	53.54 101	63.91 79	73.74 113	75.27 90	84.45 138	88.98 78
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
v144192		62.05 143	61.46 148	62.73 136	66.59 169	79.87 116	69.30 126	55.88 141	41.50 175	39.41 146	37.23 170	36.45 180	59.62 120	72.69 136	73.51 130	85.61 104	88.93 79

LGP-MVS_train		72.02 63	73.18 64	70.67 65	82.13 55	80.26 114	79.58 46	63.04 71	70.09 56	51.98 76	65.06 43	55.62 94	62.49 87	75.97 87	76.32 78	84.80 130	88.93 79

conf0.002		67.12 89	67.13 95	67.11 85	77.95 77	82.11 80	71.71 92	63.06 69	49.16 120	43.43 114	47.76 94	48.79 113	61.42 92	76.61 76	76.55 75	85.07 115	88.92 81

v2v482		63.68 117	62.85 129	64.65 104	68.01 154	80.46 111	71.90 84	57.60 122	44.26 150	42.82 132	39.80 143	38.62 170	61.56 91	73.06 118	74.86 94	86.03 72	88.90 82

conf0.01		66.60 90	66.18 101	67.09 86	77.90 78	82.02 81	71.71 92	63.05 70	49.16 120	43.41 116	46.23 109	45.78 121	61.42 92	76.55 78	74.63 97	85.04 116	88.87 83

v1		63.49 119	62.77 132	64.32 110	68.13 148	80.70 105	71.70 98	57.43 126	43.69 156	42.89 128	39.03 151	39.77 158	59.93 116	72.93 123	74.48 108	85.86 87	88.77 84

v1192		62.25 138	61.64 146	62.96 130	66.88 166	79.72 118	69.96 121	55.77 143	41.58 174	39.42 145	37.05 172	35.96 185	60.50 109	74.30 106	74.09 123	85.24 109	88.76 85

v1141		63.48 120	62.75 134	64.32 110	68.13 148	80.69 106	71.69 99	57.43 126	43.66 158	42.83 131	39.02 152	39.74 160	59.95 114	72.94 122	74.49 106	85.86 87	88.75 86

divwei89l23v2f112		63.48 120	62.76 133	64.32 110	68.13 148	80.68 107	71.71 92	57.43 126	43.69 156	42.84 129	39.01 153	39.75 159	59.94 115	72.93 123	74.49 106	85.86 87	88.75 86

v6		64.09 109	63.40 119	64.90 96	68.28 144	80.78 100	71.85 86	57.64 121	46.73 136	45.18 103	39.40 145	40.89 143	60.54 107	72.86 126	74.40 111	85.92 79	88.72 88

v1neww		64.08 110	63.38 120	64.89 98	68.27 146	80.77 102	71.84 87	57.65 119	46.66 138	45.10 104	39.40 145	40.86 144	60.57 104	72.86 126	74.40 111	85.92 79	88.71 89

v7new		64.08 110	63.38 120	64.89 98	68.27 146	80.77 102	71.84 87	57.65 119	46.66 138	45.10 104	39.40 145	40.86 144	60.57 104	72.86 126	74.40 111	85.92 79	88.71 89

V42		62.86 133	62.97 127	62.74 135	60.84 192	78.99 126	71.46 103	57.13 133	46.85 134	44.28 110	38.87 154	40.73 148	57.63 137	72.60 138	74.14 121	85.09 113	88.63 91

thres400		65.18 102	64.44 113	66.04 91	76.40 92	82.63 75	71.52 102	64.27 60	44.93 149	40.69 142	41.86 129	40.79 146	58.12 133	77.67 67	74.64 96	85.26 108	88.56 92

tfpn111		66.52 91	66.12 102	66.98 88	77.70 79	81.58 87	71.71 92	62.94 75	49.16 120	43.28 119	51.38 78	41.34 132	61.42 92	76.24 80	74.63 97	84.84 124	88.52 93

conf200view11		65.89 96	64.96 108	66.98 88	77.70 79	81.58 87	71.71 92	62.94 75	49.16 120	43.28 119	43.24 115	41.34 132	61.42 92	76.24 80	74.63 97	84.84 124	88.52 93

tfpn200view9		65.90 95	64.96 108	67.00 87	77.70 79	81.58 87	71.71 92	62.94 75	49.16 120	43.40 117	43.24 115	41.34 132	61.42 92	76.24 80	74.63 97	84.84 124	88.52 93

v1921920		61.66 150	61.10 151	62.31 139	66.32 170	79.57 120	68.41 131	55.49 147	41.03 176	38.69 152	36.64 178	35.27 191	59.60 121	73.23 116	73.41 132	85.37 106	88.51 96

tpmp4_e23		69.38 72	69.47 80	69.28 71	78.20 75	82.35 79	75.92 60	49.20 188	64.15 73	59.96 59	47.93 90	55.77 92	68.06 61	73.05 120	74.53 103	84.34 140	88.50 97

PMMVS		70.37 70	75.06 56	64.90 96	71.46 122	81.88 82	64.10 153	55.64 145	71.31 53	46.69 95	70.69 33	58.56 77	69.53 52	79.03 59	75.63 86	81.96 173	88.32 98

OPM-MVS		72.74 59	70.93 74	74.85 48	85.30 45	84.34 67	82.82 27	69.79 27	49.96 115	55.39 72	54.09 71	60.14 75	70.04 49	80.38 52	79.43 55	85.74 94	88.20 99

v1240		61.09 154	60.55 156	61.72 145	65.92 174	79.28 124	67.16 139	54.91 153	39.79 181	38.10 155	36.08 180	34.64 192	59.15 128	72.86 126	73.36 134	85.10 111	87.84 100

Fast-Effi-MVS+-dtu		63.05 127	64.72 112	61.11 148	71.21 126	76.81 150	70.72 115	43.13 210	52.51 110	35.34 172	46.55 107	46.36 118	61.40 97	71.57 152	71.44 158	84.84 124	87.79 101

MAR-MVS		77.19 43	78.37 41	75.81 41	89.87 16	90.58 17	79.33 48	65.56 53	77.62 44	58.33 62	59.24 59	67.98 49	74.83 22	82.37 33	83.12 25	86.95 52	87.67 102
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
view600		63.91 115	63.27 122	64.66 103	75.57 98	81.73 83	69.71 123	63.04 71	43.97 152	39.18 148	41.09 133	40.24 154	55.38 143	76.28 79	72.04 152	85.08 114	87.52 103

thres600view7		63.77 116	63.14 124	64.51 105	75.49 99	81.61 85	69.59 124	62.95 73	43.96 153	38.90 150	41.09 133	40.24 154	55.25 145	76.24 80	71.54 155	84.89 121	87.30 104

FC-MVSNet-train		68.83 76	68.29 87	69.47 69	78.35 74	79.94 115	64.72 150	66.38 46	54.96 102	54.51 73	56.75 62	47.91 116	66.91 69	75.57 92	75.75 84	85.92 79	87.12 105

v148		62.00 144	61.19 150	62.96 130	67.46 164	79.49 121	67.87 132	57.66 118	42.30 169	45.02 106	38.20 164	38.89 169	54.77 146	69.83 169	72.60 148	84.96 117	87.01 106

CHOSEN 280x420		62.23 140	66.57 98	57.17 174	59.88 196	68.92 193	61.20 173	42.28 212	54.17 106	39.57 144	47.78 93	64.97 57	62.68 85	73.85 110	69.52 175	77.43 200	86.75 107

Fast-Effi-MVS+		67.59 81	67.56 92	67.62 82	73.67 108	81.14 96	71.12 108	54.79 156	58.88 83	50.61 84	46.70 106	47.05 117	69.12 56	76.06 86	76.44 76	86.43 60	86.65 108

GA-MVS		64.55 106	65.76 106	63.12 129	69.68 132	81.56 90	69.59 124	58.16 113	45.23 147	35.58 171	47.01 104	41.82 131	59.41 124	79.62 55	78.54 61	86.32 61	86.56 109

AdaColmap		76.23 47	73.55 61	79.35 20	89.38 22	85.00 64	79.99 45	73.04 15	76.60 46	71.17 22	55.18 64	57.99 84	77.87 10	76.82 75	76.82 73	84.67 133	86.45 110

view800		63.02 128	62.69 135	63.39 125	74.79 104	80.76 104	67.83 133	61.93 84	43.16 163	37.78 159	40.43 138	39.73 161	53.16 150	75.01 94	73.32 135	84.87 123	86.43 111

Effi-MVS+-dtu		64.58 105	64.08 114	65.16 94	73.04 113	75.17 161	70.68 116	56.23 139	54.12 107	44.71 108	47.42 96	51.10 108	63.82 80	68.08 177	66.32 190	82.47 168	86.38 112

CDS-MVSNet		64.22 108	65.89 105	62.28 140	70.05 130	80.59 109	69.91 122	57.98 115	43.53 159	46.58 96	48.22 89	50.76 109	46.45 174	75.68 89	76.08 80	82.70 164	86.34 113
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
OMC-MVS		74.03 53	75.82 54	71.95 60	79.56 68	80.98 97	75.35 69	63.21 68	84.48 20	61.83 47	61.54 52	66.89 52	69.41 54	76.60 77	74.07 124	82.34 170	86.15 114

v10		63.00 129	62.22 139	63.90 118	67.88 158	77.78 138	71.59 101	54.34 158	45.37 145	42.76 133	38.53 158	38.93 167	61.05 99	74.39 101	74.52 104	85.75 91	86.04 115

IterMVS-LS		66.08 94	66.56 99	65.51 93	73.67 108	74.88 162	70.89 114	53.55 164	50.42 113	48.32 90	50.59 81	55.66 93	61.83 89	73.93 108	74.42 110	84.82 129	86.01 116
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
Vis-MVSNet		65.53 99	69.83 79	60.52 150	70.80 129	84.59 66	66.37 146	55.47 148	48.40 128	40.62 143	57.67 60	58.43 80	45.37 181	77.49 68	76.24 79	84.47 137	85.99 117
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
UniMVSNet (Re)		60.62 157	62.93 128	57.92 166	67.64 162	77.90 136	61.75 170	61.24 91	49.83 116	29.80 190	42.57 122	40.62 152	43.36 184	70.49 162	73.27 138	83.76 149	85.81 118

v8		63.44 123	62.58 136	64.43 107	68.28 144	78.07 134	71.82 89	54.85 154	46.70 137	45.20 102	39.40 145	40.91 142	60.54 107	72.85 130	74.39 116	85.92 79	85.76 119

CNLPA		71.37 66	70.27 77	72.66 57	80.79 65	81.33 93	71.07 111	65.75 51	82.36 26	64.80 40	42.46 125	56.49 87	72.70 34	73.00 121	70.52 170	80.84 181	85.76 119

anonymousdsp		54.99 184	57.24 180	52.36 191	53.82 213	71.75 187	51.49 195	48.14 192	33.74 204	33.66 179	38.34 161	36.13 183	47.54 168	64.53 189	70.60 169	79.53 192	85.59 121

tfpn		62.54 134	62.79 131	62.25 141	74.16 106	79.86 117	66.07 147	60.97 95	42.43 168	36.41 163	39.88 142	43.76 127	51.25 157	73.85 110	74.17 120	84.67 133	85.57 122

UniMVSNet_NR-MVSNet		62.30 137	63.51 118	60.89 149	69.48 136	77.83 137	64.07 155	63.94 62	50.03 114	31.17 186	44.82 110	41.12 138	51.37 154	71.02 154	74.81 95	85.30 107	84.95 123

DU-MVS		60.87 156	61.82 143	59.76 155	66.69 167	75.87 155	64.07 155	61.96 82	49.31 118	31.17 186	42.76 119	36.95 177	51.37 154	69.67 170	73.20 141	83.30 156	84.95 123

NR-MVSNet		61.08 155	62.09 141	59.90 153	71.96 121	75.87 155	63.60 161	61.96 82	49.31 118	27.95 195	42.76 119	33.85 197	48.82 163	74.35 103	74.05 125	85.13 110	84.45 125

RPMNet		58.63 172	62.80 130	53.76 190	67.59 163	71.29 188	54.60 192	38.13 223	55.83 96	35.70 170	41.58 131	53.04 103	47.89 166	66.10 181	67.38 183	78.65 198	84.40 126

FMVSNet3		70.41 69	71.89 68	68.68 75	70.89 128	79.42 123	75.63 63	60.97 95	65.32 68	51.06 80	47.37 97	62.05 65	64.90 75	82.49 29	82.27 30	88.64 17	84.34 127

TSAR-MVS + COLMAP		73.09 56	76.86 48	68.71 74	74.97 102	82.49 78	74.51 74	61.83 85	83.16 23	49.31 88	82.22 16	51.62 107	68.94 57	78.76 62	75.52 89	82.67 165	84.23 128

IterMVS		61.87 146	63.55 117	59.90 153	67.29 165	72.20 183	67.34 138	48.56 191	47.48 132	37.86 158	47.07 102	48.27 114	54.08 148	72.12 145	73.71 128	84.30 141	83.99 129
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
conf0.05thres1000		60.33 160	59.42 170	61.40 147	73.15 112	78.25 133	65.29 149	60.30 102	36.61 192	35.75 169	33.25 188	39.23 165	50.35 160	72.18 144	72.67 147	83.57 152	83.74 130

GBi-Net		69.21 73	70.40 75	67.81 80	69.49 133	78.65 128	74.54 70	60.97 95	65.32 68	51.06 80	47.37 97	62.05 65	63.43 81	77.49 68	78.22 65	87.37 40	83.73 131

test1		69.21 73	70.40 75	67.81 80	69.49 133	78.65 128	74.54 70	60.97 95	65.32 68	51.06 80	47.37 97	62.05 65	63.43 81	77.49 68	78.22 65	87.37 40	83.73 131

FMVSNet2		68.06 80	68.57 85	67.45 83	69.49 133	78.65 128	74.54 70	60.23 105	56.29 94	49.64 87	42.13 128	57.08 86	63.43 81	81.15 46	80.99 47	87.37 40	83.73 131

v18		63.31 124	62.02 142	64.81 100	68.48 140	73.38 171	72.14 81	54.28 159	48.99 127	47.21 93	39.56 144	41.20 136	60.80 101	72.89 125	74.46 109	85.96 78	83.64 134

Vis-MVSNet (Re-imp)		62.25 138	68.74 84	54.68 184	73.70 107	78.74 127	56.51 190	57.49 124	55.22 100	26.86 198	54.56 67	61.35 71	31.06 200	73.10 117	74.90 93	82.49 167	83.31 135

ACMM	66.70 10	70.42 67	68.49 86	72.67 56	82.85 49	77.76 139	77.70 54	64.76 58	64.61 72	60.74 56	49.29 85	53.97 100	65.86 71	74.97 95	75.57 88	84.13 146	83.29 136
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
v17		62.99 131	61.70 145	64.51 105	68.40 142	73.28 173	71.80 90	54.11 161	47.87 130	46.14 97	39.29 150	41.01 140	60.60 103	72.81 132	74.39 116	85.99 76	83.25 137

v16		63.12 126	61.78 144	64.68 102	68.45 141	73.29 172	71.86 85	54.12 160	48.36 129	47.00 94	39.30 149	41.01 140	60.67 102	72.83 131	74.40 111	86.01 73	83.24 138

CR-MVSNet		62.31 136	64.75 110	59.47 157	68.63 139	71.29 188	67.53 135	43.18 208	55.83 96	41.40 136	41.04 135	55.85 90	57.29 138	72.76 133	73.27 138	78.77 196	83.23 139

PatchT		60.46 158	63.85 115	56.51 176	65.95 173	75.68 159	47.34 203	41.39 215	53.89 108	41.40 136	37.84 168	50.30 111	57.29 138	72.76 133	73.27 138	85.67 99	83.23 139

tfpnnormal		58.97 165	56.48 183	61.89 143	71.27 125	76.21 154	66.65 143	61.76 88	32.90 206	36.41 163	27.83 204	29.14 209	50.64 159	73.06 118	73.05 142	84.58 136	83.15 141

Baseline_NR-MVSNet		59.47 163	60.28 159	58.54 163	66.69 167	73.90 168	61.63 171	62.90 78	49.15 126	26.87 197	35.18 185	37.62 173	48.20 165	69.67 170	73.61 129	84.92 118	82.82 142

v15		62.07 142	60.70 153	63.67 120	68.09 151	73.00 174	71.27 104	53.41 165	43.70 155	43.43 114	38.77 155	39.83 156	59.87 117	72.74 135	74.25 118	85.98 77	82.61 143

pmmvs4		63.14 125	62.46 137	63.94 117	66.03 172	76.40 152	66.82 141	57.60 122	56.74 90	50.26 86	40.81 137	37.51 174	59.26 127	71.75 151	71.48 157	83.68 151	82.53 144

TranMVSNet+NR-MVSNet		60.38 159	61.30 149	59.30 158	68.34 143	75.57 160	63.38 164	63.78 65	46.74 135	27.73 196	42.56 123	36.84 178	47.66 167	70.36 163	74.59 101	84.91 120	82.46 145

pmmvs5		59.72 161	60.24 161	59.11 160	62.77 186	77.33 144	63.17 165	54.00 162	40.21 179	37.23 160	40.41 139	35.99 184	51.75 153	72.55 139	72.74 146	85.72 97	82.45 146

v11		61.74 148	60.47 157	63.22 128	67.83 160	72.72 180	70.31 119	52.95 174	42.75 166	41.89 135	38.16 165	38.49 171	60.40 111	74.35 103	74.40 111	85.92 79	82.39 147

V14		61.96 145	60.56 155	63.59 121	68.06 152	72.93 177	71.10 109	53.33 167	43.47 160	43.28 119	38.59 156	39.78 157	59.76 119	72.65 137	74.19 119	86.01 73	82.32 148

v7n		57.04 179	56.64 182	57.52 171	62.85 185	74.75 164	61.76 169	51.80 178	35.58 199	36.02 168	32.33 192	33.61 198	50.16 161	67.73 178	70.34 172	82.51 166	82.12 149

V9		61.85 147	60.42 158	63.51 122	68.02 153	72.85 178	70.91 113	53.24 168	43.25 162	43.27 123	38.41 160	39.73 161	59.60 121	72.55 139	74.13 122	86.04 71	82.04 150

PatchmatchNet		65.43 100	67.71 91	62.78 133	73.49 110	82.83 74	66.42 145	45.40 201	60.40 81	45.27 101	49.22 86	57.60 85	60.01 113	70.61 157	71.38 163	86.08 69	81.91 151
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
pm-mvs1		59.21 164	59.58 169	58.77 162	67.97 156	77.07 149	64.12 152	57.20 131	34.73 200	36.86 161	35.34 183	40.54 153	43.34 185	74.32 105	73.30 137	83.13 162	81.77 152

v12		61.70 149	60.27 160	63.38 126	68.00 155	72.76 179	70.63 117	53.14 170	43.01 164	42.95 127	38.25 162	39.64 163	59.48 123	72.47 141	74.05 125	86.06 70	81.71 153

FMVSNet1		63.48 120	63.07 125	63.97 116	65.31 177	76.37 153	71.77 91	57.90 116	43.32 161	45.66 99	35.06 186	49.43 112	58.57 131	77.49 68	78.22 65	84.59 135	81.60 154

v13		61.60 151	60.13 163	63.31 127	67.95 157	72.67 181	70.51 118	53.05 171	42.80 165	42.96 124	38.10 167	39.57 164	59.31 126	72.36 142	73.98 127	86.10 67	81.40 155

TAPA-MVS	67.10 9	71.45 65	73.47 63	69.10 72	77.04 86	80.78 100	73.81 77	62.10 81	80.80 33	51.28 79	60.91 54	63.80 62	67.98 62	74.59 98	72.42 149	82.37 169	80.97 156
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MDTV_nov1_ep13		65.21 101	67.28 94	62.79 132	70.91 127	81.72 84	69.28 127	49.50 185	58.08 86	43.94 111	50.50 82	56.02 89	58.86 130	70.72 156	73.37 133	84.24 142	80.52 157

v52		54.79 187	55.15 185	54.36 188	54.07 211	72.13 184	59.84 180	49.39 186	34.50 201	35.08 174	31.63 196	35.74 187	47.21 172	63.90 192	67.92 179	80.59 184	80.23 158

V4		54.78 188	55.14 186	54.37 187	54.07 211	72.13 184	59.83 181	49.39 186	34.46 203	35.11 173	31.64 195	35.72 188	47.22 171	63.90 192	67.92 179	80.59 184	80.23 158

v748		55.19 181	54.63 187	55.85 179	61.44 191	72.97 176	58.72 185	51.62 179	34.48 202	36.39 165	32.09 193	33.05 199	45.48 179	61.85 197	67.87 181	81.45 177	80.08 160

MS-PatchMatch		70.34 71	69.00 82	71.91 61	85.20 46	85.35 61	77.84 53	61.77 87	58.01 87	55.40 71	41.26 132	58.34 81	61.69 90	81.70 41	78.29 64	89.56 7	80.02 161

ACMH	59.42 14	61.59 152	59.22 173	64.36 109	78.92 73	78.26 132	67.65 134	67.48 41	39.81 180	30.98 188	38.25 162	34.59 193	61.37 98	70.55 160	73.47 131	79.74 190	79.59 162
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
EPNet_dtu		66.17 93	70.13 78	61.54 146	81.04 61	77.39 143	68.87 129	62.50 80	69.78 57	33.51 180	63.77 45	56.22 88	37.65 197	72.20 143	72.18 151	85.69 98	79.38 163
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
EPMVS		66.21 92	67.49 93	64.73 101	75.81 96	84.20 70	68.94 128	44.37 206	61.55 77	48.07 91	49.21 87	54.87 97	62.88 84	71.82 149	71.40 160	88.28 26	79.37 164

dps		64.08 110	63.22 123	65.08 95	75.27 100	79.65 119	66.68 142	46.63 200	56.94 89	55.67 70	43.96 111	43.63 128	64.00 78	69.50 172	69.82 173	82.25 171	79.02 165

tfpn_ndepth		62.95 132	63.75 116	62.02 142	76.89 88	79.48 122	64.09 154	60.98 94	49.48 117	38.73 151	49.92 84	44.79 123	47.37 169	71.91 148	71.66 154	84.07 147	79.00 166

TAMVS		58.86 167	60.91 152	56.47 177	62.38 188	77.57 140	58.97 184	52.98 172	38.76 186	36.17 166	42.26 127	47.94 115	46.45 174	70.23 164	70.79 167	81.86 174	78.82 167

ACMH+	60.36 13	61.16 153	58.38 175	64.42 108	77.37 85	74.35 167	68.45 130	62.81 79	45.86 143	38.48 153	35.71 181	37.35 175	59.81 118	67.24 179	69.80 174	79.58 191	78.32 168

pmmvs6		54.20 191	53.54 192	54.97 182	63.22 184	72.98 175	60.17 178	52.32 177	26.77 219	34.30 177	23.29 214	36.23 182	40.33 193	68.77 175	68.76 177	79.47 193	78.00 169

PLC	64.00 12	68.54 77	66.66 97	70.74 64	80.28 67	74.88 162	72.64 80	63.70 66	69.26 60	55.71 69	47.24 100	55.31 95	70.42 46	72.05 147	70.67 168	81.66 175	77.19 170
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
EG-PatchMatch MVS		58.73 169	58.03 178	59.55 156	72.32 118	80.49 110	63.44 163	55.55 146	32.49 207	38.31 154	28.87 202	37.22 176	42.84 186	74.30 106	75.70 85	84.84 124	77.14 171

thresconf0.02		63.92 114	65.18 107	62.46 137	75.91 95	80.65 108	67.51 137	63.86 63	45.00 148	33.32 181	51.38 78	51.68 106	48.34 164	75.49 93	75.13 91	85.84 90	76.91 172

pmmvs-eth3d		55.20 180	53.95 191	56.65 175	57.34 205	67.77 195	57.54 187	53.74 163	40.93 177	41.09 140	31.19 198	29.10 210	49.07 162	65.54 182	67.28 184	81.14 179	75.81 173

Anonymous20231206		52.23 195	52.80 196	51.56 193	64.70 180	69.41 191	51.01 196	58.60 111	36.63 191	22.44 205	21.80 217	31.42 204	30.52 201	66.79 180	67.83 182	82.10 172	75.73 174

CP-MVSNet		50.57 198	52.60 198	48.21 202	58.77 201	65.82 201	48.17 201	56.29 138	37.41 188	16.59 214	37.14 171	31.95 202	29.21 204	56.60 208	63.71 199	80.22 186	75.56 175

tfpn1000		58.35 174	59.96 165	56.47 177	72.78 117	77.51 141	56.66 189	59.16 107	43.74 154	29.76 191	42.79 118	42.49 129	37.04 198	68.92 174	68.98 176	83.45 154	75.25 176

PS-CasMVS		50.17 199	52.02 199	48.02 203	58.60 202	65.54 202	48.04 202	56.19 140	36.42 194	16.42 216	35.68 182	31.33 205	28.85 206	56.42 210	63.54 200	80.01 187	75.18 177

CVMVSNet		54.92 186	58.16 176	51.13 195	62.61 187	68.44 194	55.45 191	52.38 176	42.28 170	21.45 206	47.10 101	46.10 119	37.96 196	64.42 190	63.81 198	76.92 203	75.01 178

TransMVSNet (Re)		57.83 175	56.90 181	58.91 161	72.26 119	74.69 165	63.57 162	61.42 90	32.30 208	32.65 183	33.97 187	35.96 185	39.17 195	73.84 112	72.84 145	84.37 139	74.69 179

PM-MVS		50.11 200	50.38 203	49.80 197	47.23 225	62.08 211	50.91 197	44.84 205	41.90 172	36.10 167	35.22 184	26.05 217	46.83 173	57.64 204	55.42 221	72.90 212	74.32 180

PatchMatch-RL		62.22 141	60.69 154	64.01 115	68.74 138	75.75 158	59.27 182	60.35 101	56.09 95	53.80 74	47.06 103	36.45 180	64.80 76	68.22 176	67.22 185	77.10 201	74.02 181

PEN-MVS		51.04 196	52.94 194	48.82 199	61.45 190	66.00 200	48.68 200	57.20 131	36.87 190	15.36 217	36.98 173	32.72 200	28.77 207	57.63 205	66.37 189	81.44 178	74.00 182

WR-MVS		51.02 197	54.56 189	46.90 205	63.84 182	69.23 192	44.78 210	56.38 137	38.19 187	14.19 219	37.38 169	36.82 179	22.39 217	60.14 200	66.20 192	79.81 189	73.95 183

MIMVSNet		57.78 176	59.71 167	55.53 181	54.79 207	77.10 148	63.89 159	45.02 203	46.59 140	36.79 162	28.36 203	40.77 147	45.84 178	74.97 95	76.58 74	86.87 54	73.60 184

USDC		59.69 162	60.03 164	59.28 159	64.04 181	71.84 186	63.15 166	55.36 149	54.90 103	35.02 175	48.34 88	29.79 208	58.16 132	70.60 158	71.33 164	79.99 188	73.42 185

MSDG		65.57 98	61.57 147	70.24 66	82.02 56	76.47 151	74.46 76	68.73 36	56.52 92	50.33 85	38.47 159	41.10 139	62.42 88	72.12 145	72.94 143	83.47 153	73.37 186

MDTV_nov1_ep13_2view		54.47 190	54.61 188	54.30 189	60.50 193	73.82 169	57.92 186	43.38 207	39.43 185	32.51 184	33.23 189	34.05 195	47.26 170	62.36 195	66.21 191	84.24 142	73.19 187

tfpnview11		58.92 166	59.60 168	58.13 164	72.99 114	77.11 147	60.48 174	60.37 100	42.10 171	29.10 192	43.45 112	40.72 149	41.67 189	70.53 161	70.43 171	84.17 145	72.85 188

WR-MVS_H		49.62 202	52.63 197	46.11 208	58.80 200	67.58 196	46.14 208	54.94 151	36.51 193	13.63 222	36.75 176	35.67 189	22.10 218	56.43 209	62.76 202	81.06 180	72.73 189

ADS-MVSNet		58.40 173	59.16 174	57.52 171	65.80 175	74.57 166	60.26 177	40.17 220	50.51 112	38.01 156	40.11 141	44.72 124	59.36 125	64.91 185	66.55 188	81.53 176	72.72 190

tfpn_n400		58.64 170	59.27 171	57.89 167	72.83 115	77.26 145	60.35 175	60.29 103	39.77 182	29.10 192	43.45 112	40.72 149	41.61 190	70.06 166	71.39 161	83.17 159	72.26 191

tfpnconf		58.64 170	59.27 171	57.89 167	72.83 115	77.26 145	60.35 175	60.29 103	39.77 182	29.10 192	43.45 112	40.72 149	41.61 190	70.06 166	71.39 161	83.17 159	72.26 191

LTVRE_ROB	47.26 16	49.41 203	49.91 205	48.82 199	64.76 179	69.79 190	49.05 198	47.12 197	20.36 228	16.52 215	36.65 177	26.96 212	50.76 158	60.47 199	63.16 201	64.73 222	72.00 193
Andreas Kuhn, Heiko HirschmÃ¼ller, Daniel Scharstein, Helmut Mayer: A TV Prior for High-Quality Scalable Multi-View Stereo Reconstruction. International Journal of Computer Vision 2016
UA-Net		64.62 104	68.23 89	60.42 151	77.53 83	81.38 92	60.08 179	57.47 125	47.01 133	44.75 107	60.68 55	71.32 40	41.84 188	73.27 115	72.25 150	80.83 182	71.68 194

test0.0.03 1		57.35 178	59.89 166	54.38 186	71.37 123	73.45 170	52.71 194	61.03 92	46.11 142	26.33 199	41.73 130	44.08 125	29.72 203	71.43 153	70.90 166	85.10 111	71.56 195

LS3D		64.54 107	62.14 140	67.34 84	80.85 63	75.79 157	69.99 120	65.87 50	60.77 79	44.35 109	42.43 126	45.95 120	65.01 73	69.88 168	68.69 178	77.97 199	71.43 196

MVS-HIRNet		53.86 192	53.02 193	54.85 183	60.30 195	72.36 182	44.63 211	42.20 213	39.45 184	43.47 113	21.66 218	34.00 196	55.47 142	65.42 183	67.16 186	83.02 163	71.08 197

DTE-MVSNet		49.82 201	51.92 200	47.37 204	61.75 189	64.38 205	45.89 209	57.33 130	36.11 195	12.79 224	36.87 174	31.93 203	25.73 212	58.01 202	65.22 194	80.75 183	70.93 198

testpf		43.39 213	47.17 210	38.98 217	65.58 176	47.38 227	36.09 220	31.67 230	36.97 189	19.47 209	33.01 190	35.62 190	23.61 216	50.86 222	56.08 218	57.48 227	70.27 199

EU-MVSNet		44.84 211	47.85 208	41.32 215	49.26 220	56.59 218	43.07 212	47.64 196	33.03 205	13.82 220	36.78 175	30.99 206	24.37 215	53.80 217	55.57 220	69.78 217	68.21 200

SixPastTwentyTwo		49.11 204	49.22 206	48.99 198	58.54 203	64.14 206	47.18 204	47.75 194	31.15 210	24.42 201	41.01 136	26.55 213	44.04 183	54.76 216	58.70 211	71.99 215	68.21 200

CMPMVS	43.63 17	57.67 177	55.43 184	60.28 152	72.01 120	79.00 125	62.77 167	53.23 169	41.77 173	45.42 100	30.74 199	39.03 166	53.01 151	64.81 187	64.65 196	75.26 207	68.03 202
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
ambc			42.30 216		50.36 217	49.51 224	35.47 221		32.04 209	23.53 202	17.36 223	8.95 235	29.06 205	64.88 186	56.26 217	61.29 224	67.12 203

FMVSNet5		58.86 167	60.24 161	57.25 173	52.66 216	66.25 199	63.77 160	52.86 175	57.85 88	37.92 157	36.12 179	52.22 105	51.37 154	70.88 155	71.43 159	84.92 118	66.91 204

test2356		46.29 210	47.37 209	45.03 210	54.38 209	57.99 216	42.03 213	50.32 182	30.78 211	16.65 213	27.40 206	23.70 220	29.86 202	61.20 198	64.31 197	76.93 202	66.22 205

TDRefinement		52.70 193	51.02 201	54.66 185	57.41 204	65.06 203	61.47 172	54.94 151	44.03 151	33.93 178	30.13 201	27.57 211	46.17 176	61.86 196	62.48 205	74.01 211	66.06 206

RPSCF		55.07 183	58.06 177	51.57 192	48.87 223	58.95 213	53.68 193	41.26 218	62.42 74	45.88 98	54.38 69	54.26 98	53.75 149	57.15 206	53.53 222	66.01 221	65.75 207

pmmvs3		41.86 216	42.29 217	41.36 213	39.80 226	52.66 221	38.93 219	35.85 229	23.40 223	20.22 208	19.30 219	20.84 224	40.56 192	55.98 211	58.79 210	72.80 213	65.03 208

FC-MVSNet-test		47.24 208	54.37 190	38.93 218	59.49 198	58.25 215	34.48 222	53.36 166	45.66 144	6.66 233	50.62 80	42.02 130	16.62 227	58.39 201	61.21 207	62.99 223	64.40 209

TinyColmap		52.66 194	50.09 204	55.65 180	59.72 197	64.02 207	57.15 188	52.96 173	40.28 178	32.51 184	32.42 191	20.97 223	56.65 140	63.95 191	65.15 195	74.91 208	63.87 210

Anonymous20231211		40.44 218	39.25 219	41.84 212	54.29 210	57.29 217	41.10 215	49.06 189	17.67 231	10.15 228	10.63 230	16.79 229	25.15 214	52.14 218	56.70 216	71.30 216	63.51 211

N_pmnet		47.67 207	47.00 211	48.45 201	54.72 208	62.78 209	46.95 205	51.25 180	36.01 196	26.09 200	26.59 208	25.93 218	35.50 199	55.67 212	59.01 209	76.22 204	63.04 212

COLMAP_ROB	51.17 15	55.13 182	52.90 195	57.73 170	73.47 111	67.21 197	62.13 168	55.82 142	47.83 131	34.39 176	31.60 197	34.24 194	44.90 182	63.88 194	62.52 204	75.67 205	63.02 213
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
test20.03		47.23 209	48.69 207	45.53 209	63.28 183	64.39 204	41.01 216	56.93 134	29.16 213	15.21 218	23.90 210	30.76 207	17.51 226	64.63 188	65.26 193	79.21 195	62.71 214

testus		42.30 214	43.69 213	40.67 216	53.21 214	53.50 220	31.81 224	49.96 183	27.06 217	11.55 226	25.67 209	19.00 226	25.20 213	55.34 213	62.59 203	72.31 214	62.69 215

new-patchmatchnet		42.21 215	42.97 215	41.33 214	53.05 215	59.89 212	39.38 217	49.61 184	28.26 216	12.10 225	22.17 216	21.54 222	19.22 222	50.96 221	56.04 219	74.61 210	61.92 216

MDA-MVSNet-bldmvs		44.15 212	42.27 218	46.34 206	38.34 228	62.31 210	46.28 206	55.74 144	29.83 212	20.98 207	27.11 207	16.45 230	41.98 187	41.11 228	57.47 213	74.72 209	61.65 217

testgi		48.51 205	50.53 202	46.16 207	64.78 178	67.15 198	41.54 214	54.81 155	29.12 214	17.03 212	32.07 194	31.98 201	20.15 221	65.26 184	67.00 187	78.67 197	61.10 218

LP		48.21 206	46.65 212	50.03 196	60.39 194	63.86 208	48.73 199	38.71 222	35.60 198	32.99 182	23.31 213	24.95 219	40.07 194	57.73 203	61.56 206	79.29 194	59.51 219

testmv		37.40 221	37.95 221	36.76 221	48.97 221	49.33 225	28.65 228	46.74 198	18.34 229	7.68 231	16.80 226	14.47 231	19.18 223	51.72 219	56.93 214	69.36 218	58.09 220

test1235678		37.40 221	37.94 222	36.76 221	48.97 221	49.30 226	28.65 228	46.73 199	18.33 230	7.68 231	16.79 227	14.46 232	19.18 223	51.72 219	56.92 215	69.36 218	58.07 221

MIMVSNet1		40.84 217	43.46 214	37.79 220	32.14 230	58.92 214	39.24 218	50.83 181	27.00 218	11.29 227	16.76 228	26.53 214	17.75 225	57.14 207	61.12 208	75.46 206	56.78 222

new_pmnet		33.19 223	35.52 224	30.47 224	27.55 234	45.31 229	29.29 227	30.92 231	29.00 215	9.88 230	18.77 220	17.64 228	26.77 211	44.07 223	45.98 225	58.41 226	47.87 223

FPMVS		39.11 219	36.39 223	42.28 211	55.97 206	45.94 228	46.23 207	41.57 214	35.73 197	22.61 203	23.46 212	19.82 225	28.32 210	43.57 224	40.67 227	58.96 225	45.54 224

test12356		29.92 225	31.49 225	28.08 225	38.46 227	37.74 231	21.36 231	40.17 220	16.83 232	5.61 235	15.66 229	11.48 233	6.60 233	42.01 226	51.23 223	56.29 228	45.52 225

no-one		26.96 226	26.51 227	27.49 227	37.87 229	39.14 230	17.12 233	41.31 217	12.02 234	3.68 237	8.04 232	8.42 236	10.67 231	28.11 230	45.96 226	54.27 229	43.89 226

1111		38.93 220	38.98 220	38.86 219	50.10 218	50.42 222	29.52 225	38.00 224	22.67 224	17.99 210	17.40 221	26.26 215	28.72 208	54.86 214	58.20 212	68.82 220	43.08 227

PMVS	27.44 18	32.08 224	29.07 226	35.60 223	48.33 224	24.79 233	26.97 230	41.34 216	20.45 227	22.50 204	17.11 225	18.64 227	20.44 220	41.99 227	38.06 228	54.02 230	42.44 228
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
PMMVS2		20.45 229	22.31 230	18.27 232	20.52 235	26.73 232	14.85 235	28.43 233	13.69 233	0.79 240	10.35 231	9.10 234	3.83 235	27.64 231	32.87 229	41.17 231	35.81 229

DeepMVS_CX						19.81 236	17.01 234	10.02 234	23.61 222	5.85 234	17.21 224	8.03 237	21.13 219	22.60 232		21.42 236	30.01 230

MVE	15.98 19	14.37 232	16.36 231	12.04 234	7.72 237	20.24 235	5.90 239	29.05 232	8.28 237	3.92 236	4.72 235	2.42 239	9.57 232	18.89 233	31.46 230	16.07 237	28.53 231
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
Gipuma		24.91 228	24.61 228	25.26 229	31.47 231	21.59 234	18.06 232	37.53 226	25.43 221	10.03 229	4.18 236	4.25 238	14.85 228	43.20 225	47.03 224	39.62 232	26.55 232
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
E-PMN		15.08 230	11.65 232	19.08 230	28.73 232	12.31 237	6.95 238	36.87 228	10.71 236	3.63 238	5.13 233	2.22 241	13.81 230	11.34 234	18.50 232	24.49 234	21.32 233

EMVS		14.40 231	10.71 233	18.70 231	28.15 233	12.09 238	7.06 237	36.89 227	11.00 235	3.56 239	4.95 234	2.27 240	13.91 229	10.13 235	16.06 233	22.63 235	18.51 234

test123		0.05 233	0.08 234	0.01 235	0.00 239	0.01 240	0.01 242	0.00 238	0.05 238	0.00 242	0.16 237	0.00 243	0.04 236	0.02 237	0.05 234	0.00 240	0.26 235

.test1245		25.86 227	24.56 229	27.39 228	50.10 218	50.42 222	29.52 225	38.00 224	22.67 224	17.99 210	17.40 221	26.26 215	28.72 208	54.86 214	0.05 234	0.01 238	0.24 236

testmvs		0.05 233	0.08 234	0.01 235	0.00 239	0.01 240	0.03 241	0.01 237	0.05 238	0.00 242	0.14 238	0.01 242	0.03 238	0.05 236	0.05 234	0.01 238	0.24 236

sosnet-low-res		0.00 235	0.00 236	0.00 237	0.00 239	0.00 242	0.00 243	0.00 238	0.00 240	0.00 242	0.00 239	0.00 243	0.00 239	0.00 238	0.00 237	0.00 240	0.00 238

sosnet		0.00 235	0.00 236	0.00 237	0.00 239	0.00 242	0.00 243	0.00 238	0.00 240	0.00 242	0.00 239	0.00 243	0.00 239	0.00 238	0.00 237	0.00 240	0.00 238

MTAPA										78.32 6		79.42 18

MTMP										76.04 11		76.65 23

Patchmatch-RL test							2.17 240

tmp_tt				16.09 233	13.07 236	8.12 239	13.61 236	2.08 235	55.09 101	30.10 189	40.26 140	22.83 221	5.35 234	29.91 229	25.25 231	32.33 233

XVS					82.43 50	86.27 55	75.70 61			61.07 52		72.27 33				85.67 99

X-MVStestdata					82.43 50	86.27 55	75.70 61			61.07 52		72.27 33				85.67 99

mPP-MVS					86.96 37							70.61 43

NP-MVS									81.60 31

Patchmtry						78.06 135	67.53 135	43.18 208		41.40 136

High-resolution multi-view benchmark