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.

SMA-MVS		77.32 7	82.51 7	71.26 7	75.43 14	80.19 8	82.22 7	58.26 3	84.83 7	64.36 7	78.19 15	83.46 6	63.61 8	81.00 1	80.28 1	83.66 4	89.62 6
Yufeng Yin; Xiaoyan Liu; Zichao Zhang: SMA-MVS: Segmentation-Guided Multi-Scale Anchor Deformation Patch Multi-View Stereo. IEEE Transactions on Circuits and Systems for Video Technology
SteuartSystems-ACMMP		75.23 13	79.60 15	70.13 13	76.81 6	78.92 12	81.74 8	57.99 6	75.30 29	59.83 25	75.69 18	78.45 24	60.48 29	80.58 2	79.77 2	83.94 3	88.52 10
Skip Steuart: Steuart Systems R&D Blog.
CNVR-MVS		75.62 12	79.91 14	70.61 10	75.76 10	78.82 14	81.66 9	57.12 13	79.77 16	63.04 13	70.69 24	81.15 16	62.99 11	80.23 5	79.54 3	83.11 8	89.16 8

ACMMP_NAP		76.15 9	81.17 9	70.30 11	74.09 20	79.47 10	81.59 12	57.09 14	81.38 11	63.89 10	79.02 13	80.48 19	62.24 17	80.05 6	79.12 4	82.94 11	88.64 9

HPM-MVS++		76.01 10	80.47 12	70.81 9	76.60 8	74.96 35	80.18 17	58.36 2	81.96 10	63.50 11	78.80 14	82.53 11	64.40 6	78.74 10	78.84 5	81.81 32	87.46 18

SED-MVS		79.21 1	84.74 2	72.75 1	78.66 2	81.96 2	82.94 4	58.16 4	86.82 2	67.66 1	88.29 4	86.15 3	66.42 2	80.41 4	78.65 6	82.65 16	90.92 2

DVP-MVS		78.77 2	84.89 1	71.62 4	78.04 3	82.05 1	81.64 10	57.96 7	87.53 1	66.64 2	88.77 1	86.31 1	63.16 10	79.99 7	78.56 7	82.31 22	91.03 1
Zhenlong Yuan, Jinguo Luo, Fei Shen, Zhaoxin Li, Cong Liu, Tianlu Mao, Zhaoqi Wang: DVP-MVS: Synergize Depth-Edge and Visibility Prior for Multi-View Stereo. AAAI2025
MSP-MVS		77.82 5	83.46 5	71.24 8	75.26 16	80.22 7	82.95 3	57.85 8	85.90 3	64.79 5	88.54 3	83.43 7	66.24 3	78.21 17	78.56 7	80.34 45	89.39 7
Zhenlong Yuan, Cong Liu, Fei Shen, Zhaoxin Li, Jingguo luo, Tianlu Mao and Zhaoqi Wang: MSP-MVS: Multi-granularity Segmentation Prior Guided Multi-View Stereo. AAAI2025
DVP-MVS++		78.76 3	84.44 3	72.14 2	76.63 7	81.93 3	82.92 5	58.10 5	85.86 4	66.53 3	87.86 5	86.16 2	66.45 1	80.46 3	78.53 9	82.19 27	90.29 4

NCCC		74.27 19	77.83 24	70.13 13	75.70 11	77.41 22	80.51 15	57.09 14	78.25 20	62.28 17	65.54 37	78.26 25	62.18 18	79.13 8	78.51 10	83.01 10	87.68 17

DeepC-MVS	66.32 2	73.85 22	78.10 23	68.90 22	67.92 48	79.31 11	78.16 28	59.28 1	78.24 21	61.13 19	67.36 35	76.10 32	63.40 9	79.11 9	78.41 11	83.52 5	88.16 13
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
MP-MVS		74.31 18	78.87 18	68.99 21	73.49 23	78.56 15	79.25 22	56.51 17	75.33 27	60.69 22	75.30 19	79.12 23	61.81 20	77.78 21	77.93 12	82.18 29	88.06 14
Rongxuan Tan, Qing Wang, et al.: MP-MVS: Multi-Scale Windows PatchMatch and Planar Prior Multi-View Stereo.
HFP-MVS		74.87 15	78.86 20	70.21 12	73.99 21	77.91 18	80.36 16	56.63 16	78.41 19	64.27 8	74.54 20	77.75 27	62.96 12	78.70 12	77.82 13	83.02 9	86.91 21

ACMMPR		73.79 23	78.41 21	68.40 24	72.35 27	77.79 19	79.32 20	56.38 18	77.67 23	58.30 31	74.16 21	76.66 28	61.40 22	78.32 14	77.80 14	82.68 15	86.51 22

DPE-MVS		78.11 4	83.84 4	71.42 5	77.82 5	81.32 4	82.92 5	57.81 9	84.04 8	63.19 12	88.63 2	86.00 4	64.52 5	78.71 11	77.63 15	82.26 23	90.57 3
Kehua Chen, Zhenlong Yuan, Tianlu Mao, Zhaoqi Wang: Dual-Level Precision Edges Guided Multi-View Stereo with Accurate Planarization. AAAI2025
DeepPCF-MVS	66.49 1	74.25 20	80.97 10	66.41 31	67.75 50	78.87 13	75.61 38	54.16 33	84.86 6	58.22 32	77.94 16	81.01 17	62.52 15	78.34 13	77.38 16	80.16 48	88.40 11

X-MVS		71.18 32	75.66 32	65.96 35	71.71 29	76.96 25	77.26 32	55.88 22	72.75 37	54.48 47	64.39 43	74.47 37	54.19 65	77.84 20	77.37 17	82.21 26	85.85 26

PGM-MVS		72.89 25	77.13 26	67.94 25	72.47 26	77.25 23	79.27 21	54.63 29	73.71 34	57.95 33	72.38 22	75.33 34	60.75 27	78.25 16	77.36 18	82.57 19	85.62 28

CSCG		74.68 16	79.22 16	69.40 17	75.69 12	80.01 9	79.12 23	52.83 41	79.34 17	63.99 9	70.49 25	82.02 12	60.35 32	77.48 24	77.22 19	84.38 1	87.97 15

CP-MVS		72.63 27	76.95 27	67.59 26	70.67 35	75.53 33	77.95 30	56.01 21	75.65 26	58.82 28	69.16 29	76.48 30	60.46 30	77.66 22	77.20 20	81.65 36	86.97 20

DeepC-MVS_fast	65.08 3	72.00 29	76.11 28	67.21 28	68.93 44	77.46 21	76.54 34	54.35 31	74.92 31	58.64 30	65.18 39	74.04 42	62.62 14	77.92 19	77.02 21	82.16 30	86.21 23
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
APDe-MVS		77.58 6	82.93 6	71.35 6	77.86 4	80.55 6	83.38 1	57.61 10	85.57 5	61.11 20	86.10 7	82.98 8	64.76 4	78.29 15	76.78 22	83.40 6	90.20 5

DPM-MVS		72.80 26	75.90 29	69.19 20	75.51 13	77.68 20	81.62 11	54.83 26	75.96 25	62.06 18	63.96 46	76.58 29	58.55 40	76.66 32	76.77 23	82.60 18	83.68 40

CDPH-MVS		71.47 31	75.82 31	66.41 31	72.97 25	77.15 24	78.14 29	54.71 27	69.88 46	53.07 55	70.98 23	74.83 36	56.95 52	76.22 33	76.57 24	82.62 17	85.09 33

OPM-MVS		69.33 37	71.05 45	67.32 27	72.34 28	75.70 32	79.57 19	56.34 19	55.21 74	53.81 52	59.51 63	68.96 55	59.67 34	77.61 23	76.44 25	82.19 27	83.88 38
Ray L. Khuboni, Hongjun Xu: Octagram Propagation Matching for Multi-Scale View Stereopsis (OPM-MVS).
ACMMP		71.57 30	75.84 30	66.59 30	70.30 39	76.85 28	78.46 27	53.95 34	73.52 35	55.56 38	70.13 26	71.36 47	58.55 40	77.00 27	76.23 26	82.71 14	85.81 27
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
APD-MVS		75.80 11	80.90 11	69.86 15	75.42 15	78.48 16	81.43 13	57.44 12	80.45 14	59.32 26	85.28 8	80.82 18	63.96 7	76.89 28	76.08 27	81.58 38	88.30 12
Yuesong Wang, Zhaojie Zeng and etc.: Adaptive Patch Deformation for Textureless-Resilient Multi-View Stereo. CVPR2023
train_agg		73.89 21	78.25 22	68.80 23	75.25 17	72.27 50	79.75 18	56.05 20	74.87 32	58.97 27	81.83 11	79.76 21	61.05 25	77.39 25	76.01 28	81.71 35	85.61 29

MCST-MVS		73.67 24	77.39 25	69.33 18	76.26 9	78.19 17	78.77 25	54.54 30	75.33 27	59.99 24	67.96 31	79.23 22	62.43 16	78.00 18	75.71 29	84.02 2	87.30 19

TSAR-MVS + MP.		75.22 14	80.06 13	69.56 16	74.61 18	72.74 48	80.59 14	55.70 23	80.80 13	62.65 15	86.25 6	82.92 9	62.07 19	76.89 28	75.66 30	81.77 34	85.19 32
Zhenlong Yuan, Jiakai Cao, Zhaoqi Wang, Zhaoxin Li: TSAR-MVS: Textureless-aware Segmentation and Correlative Refinement Guided Multi-View Stereo. Pattern Recognition
SF-MVS		77.13 8	81.70 8	71.79 3	79.32 1	80.76 5	82.96 2	57.49 11	82.82 9	64.79 5	83.69 10	84.46 5	62.83 13	77.13 26	75.21 31	83.35 7	87.85 16

3Dnovator+	62.63 4	69.51 35	72.62 39	65.88 36	68.21 47	76.47 29	73.50 48	52.74 42	70.85 42	58.65 29	55.97 76	69.95 50	61.11 24	76.80 30	75.09 32	81.09 41	83.23 43

MVS_0304		69.49 36	73.96 35	64.28 44	67.92 48	76.13 31	74.90 41	47.60 68	63.29 57	54.09 51	67.44 34	76.35 31	59.53 35	75.81 37	75.03 33	81.62 37	83.70 39

ACMM	60.30 7	67.58 47	68.82 59	66.13 33	70.59 36	72.01 52	76.54 34	54.26 32	65.64 52	54.78 46	50.35 105	61.72 80	58.74 38	75.79 38	75.03 33	81.88 31	81.17 51
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
HQP-MVS		70.88 33	75.02 33	66.05 34	71.69 30	74.47 40	77.51 31	53.17 38	72.89 36	54.88 44	70.03 27	70.48 49	57.26 48	76.02 35	75.01 35	81.78 33	86.21 23

CLD-MVS		67.02 49	71.57 42	61.71 51	71.01 34	74.81 37	71.62 50	38.91 158	71.86 40	60.70 21	64.97 41	67.88 63	51.88 91	76.77 31	74.98 36	76.11 94	69.75 122
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
MVS_111021_HR		67.62 46	70.39 49	64.39 43	69.77 40	70.45 57	71.44 52	51.72 47	60.77 63	55.06 42	62.14 55	66.40 66	58.13 43	76.13 34	74.79 37	80.19 47	82.04 48

MAR-MVS		68.04 44	70.74 47	64.90 41	71.68 31	76.33 30	74.63 43	50.48 55	63.81 54	55.52 39	54.88 82	69.90 51	57.39 47	75.42 41	74.79 37	79.71 50	80.03 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
LGP-MVS_train		68.87 39	72.03 41	65.18 39	69.33 42	74.03 43	76.67 33	53.88 35	68.46 47	52.05 59	63.21 48	63.89 69	56.31 55	75.99 36	74.43 39	82.83 13	84.18 35

PHI-MVS		69.27 38	74.84 34	62.76 50	66.83 53	74.83 36	73.88 46	49.32 60	70.61 43	50.93 62	69.62 28	74.84 35	57.25 49	75.53 39	74.32 40	78.35 66	84.17 36

ACMP	61.42 5	68.72 42	71.37 43	65.64 37	69.06 43	74.45 41	75.88 37	53.30 37	68.10 48	55.74 37	61.53 58	62.29 76	56.97 51	74.70 46	74.23 41	82.88 12	84.31 34
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
TSAR-MVS + GP.		69.71 34	73.92 36	64.80 42	68.27 46	70.56 55	71.90 49	50.75 51	71.38 41	57.46 35	68.68 30	75.42 33	60.10 33	73.47 51	73.99 42	80.32 46	83.97 37

SD-MVS		74.43 17	78.87 18	69.26 19	74.39 19	73.70 44	79.06 24	55.24 25	81.04 12	62.71 14	80.18 12	82.61 10	61.70 21	75.43 40	73.92 43	82.44 21	85.22 31
Zhenlong Yuan, Jiakai Cao, Zhaoxin Li, Hao Jiang and Zhaoqi Wang: SD-MVS: Segmentation-driven Deformation Multi-View Stereo with Spherical Refinement and EM optimization. AAAI2024
TSAR-MVS + ACMM		72.56 28	79.07 17	64.96 40	73.24 24	73.16 47	78.50 26	48.80 66	79.34 17	55.32 40	85.04 9	81.49 15	58.57 39	75.06 43	73.75 44	75.35 106	85.61 29

AdaColmap		67.89 45	68.85 58	66.77 29	73.73 22	74.30 42	75.28 39	53.58 36	70.24 44	57.59 34	51.19 102	59.19 91	60.74 28	75.33 42	73.72 45	79.69 53	77.96 69

3Dnovator	60.86 6	66.99 51	70.32 50	63.11 48	66.63 54	74.52 38	71.56 51	45.76 77	67.37 50	55.00 43	54.31 87	68.19 59	58.49 42	73.97 49	73.63 46	81.22 40	80.23 54

CANet		68.77 40	73.01 37	63.83 45	68.30 45	75.19 34	73.73 47	47.90 67	63.86 53	54.84 45	67.51 33	74.36 40	57.62 44	74.22 48	73.57 47	80.56 43	82.36 45

DELS-MVS		65.87 53	70.30 51	60.71 54	64.05 70	72.68 49	70.90 53	45.43 81	57.49 69	49.05 69	64.43 42	68.66 56	55.11 63	74.31 47	73.02 48	79.70 51	81.51 49
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
EPNet		65.14 59	69.54 55	60.00 58	66.61 55	67.67 75	67.53 63	55.32 24	62.67 59	46.22 79	67.74 32	65.93 67	48.07 109	72.17 57	72.12 49	76.28 90	78.47 65
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
CPTT-MVS		68.76 41	73.01 37	63.81 46	65.42 60	73.66 45	76.39 36	52.08 43	72.61 38	50.33 64	60.73 59	72.65 45	59.43 36	73.32 52	72.12 49	79.19 59	85.99 25

CS-MVS		65.88 52	69.71 54	61.41 52	61.76 80	68.14 67	67.65 61	44.00 105	59.14 67	52.69 56	65.19 38	68.13 60	60.90 26	74.74 45	71.58 51	81.46 39	81.04 52

Effi-MVS+		63.28 62	65.96 69	60.17 57	64.26 66	68.06 69	68.78 58	45.71 79	54.08 77	46.64 76	55.92 77	63.13 73	55.94 59	70.38 75	71.43 52	79.68 54	78.70 62

QAPM		65.27 55	69.49 56	60.35 55	65.43 59	72.20 51	65.69 84	47.23 69	63.46 55	49.14 67	53.56 88	71.04 48	57.01 50	72.60 56	71.41 53	77.62 70	82.14 47

ETV-MVS		63.23 63	66.08 68	59.91 59	63.13 74	68.13 68	67.62 62	44.62 92	53.39 82	46.23 78	58.74 65	58.19 94	57.45 46	73.60 50	71.38 54	80.39 44	79.13 59

DROMVSNet		67.01 50	70.27 52	63.21 47	67.21 51	70.47 56	69.01 55	46.96 71	59.16 66	53.23 54	64.01 45	69.71 53	60.37 31	74.92 44	71.24 55	82.50 20	82.41 44

canonicalmvs		65.62 54	72.06 40	58.11 68	63.94 71	71.05 53	64.49 94	43.18 126	74.08 33	47.35 72	64.17 44	71.97 46	51.17 94	71.87 58	70.74 56	78.51 64	80.56 53

EG-PatchMatch MVS		56.98 106	58.24 130	55.50 91	64.66 63	68.62 63	61.48 105	43.63 114	38.44 191	41.44 102	38.05 181	46.18 158	43.95 127	71.71 60	70.61 57	77.87 67	74.08 102

MSLP-MVS++		68.17 43	70.72 48	65.19 38	69.41 41	70.64 54	74.99 40	45.76 77	70.20 45	60.17 23	56.42 74	73.01 43	61.14 23	72.80 54	70.54 58	79.70 51	81.42 50

IB-MVS	54.11 11	58.36 95	60.70 96	55.62 90	58.67 100	68.02 71	61.56 103	43.15 127	46.09 138	44.06 90	44.24 152	50.99 127	48.71 103	66.70 127	70.33 59	77.60 71	78.50 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
anonymousdsp		52.84 138	57.78 134	47.06 156	40.24 201	58.95 149	53.70 158	33.54 191	36.51 198	32.69 141	43.88 154	45.40 162	47.97 110	67.17 119	70.28 60	74.22 115	82.29 46

ACMH	52.42 13	58.24 97	59.56 117	56.70 84	66.34 57	69.59 58	66.71 72	49.12 61	46.08 139	28.90 159	42.67 169	41.20 186	52.60 83	71.39 62	70.28 60	76.51 86	75.72 89
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
IS_MVSNet		57.95 100	64.26 80	50.60 121	61.62 82	65.25 101	57.18 130	45.42 82	50.79 97	26.49 172	57.81 70	60.05 88	34.51 173	71.24 65	70.20 62	78.36 65	74.44 98

CS-MVS-test		65.18 57	68.70 60	61.07 53	61.92 77	68.06 69	67.09 69	45.18 85	58.47 68	52.02 60	65.76 36	66.44 65	59.24 37	72.71 55	70.05 63	80.98 42	79.40 58

PVSNet_Blended_VisFu		63.65 61	66.92 62	59.83 60	60.03 92	73.44 46	66.33 75	48.95 62	52.20 93	50.81 63	56.07 75	60.25 87	53.56 71	73.23 53	70.01 64	79.30 56	83.24 42

Vis-MVSNet		58.48 91	65.70 71	50.06 126	53.40 145	67.20 81	60.24 114	43.32 123	48.83 115	30.23 152	62.38 54	61.61 81	40.35 145	71.03 66	69.77 65	72.82 134	79.11 60
Jingyang Zhang, Yao Yao, Shiwei Li, Zixin Luo, Tian Fang: Visibility-aware Multiview Stereo Network. BMVC 2020
test2506		55.82 118	59.57 116	51.46 117	60.39 89	64.55 107	58.69 122	48.87 63	53.91 78	26.99 168	48.97 111	41.72 185	37.71 157	70.96 67	69.49 66	76.08 95	67.37 142

ECVR-MVS		56.44 113	60.74 95	51.42 118	60.39 89	64.55 107	58.69 122	48.87 63	53.91 78	26.76 170	45.55 143	53.43 115	37.71 157	70.96 67	69.49 66	76.08 95	67.32 144

Fast-Effi-MVS+		60.36 76	63.35 83	56.87 82	58.70 99	65.86 94	65.08 90	37.11 172	53.00 87	45.36 84	52.12 96	56.07 107	56.27 56	71.28 64	69.42 68	78.71 60	75.69 90

PCF-MVS	59.98 8	67.32 48	71.04 46	62.97 49	64.77 62	74.49 39	74.78 42	49.54 57	67.44 49	54.39 50	58.35 68	72.81 44	55.79 61	71.54 61	69.24 69	78.57 61	83.41 41
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
EIA-MVS		61.53 74	63.79 82	58.89 65	63.82 72	67.61 76	65.35 87	42.15 136	49.98 101	45.66 82	57.47 72	56.62 101	56.59 54	70.91 69	69.15 70	79.78 49	74.80 96

EPP-MVSNet		59.39 81	65.45 72	52.32 114	60.96 85	67.70 74	58.42 124	44.75 90	49.71 103	27.23 167	59.03 64	62.20 77	43.34 131	70.71 70	69.13 71	79.25 58	79.63 57

test1111		55.24 124	59.98 109	49.71 127	59.80 95	64.10 112	56.48 137	49.34 59	52.27 92	21.56 184	44.49 150	51.96 121	35.93 168	70.59 71	69.07 72	75.13 108	67.40 140

TranMVSNet+NR-MVSNet		55.87 116	60.14 106	50.88 120	59.46 98	63.82 113	57.93 126	52.98 39	48.94 113	20.52 187	52.87 91	47.33 145	36.81 165	69.12 85	69.03 73	77.56 73	69.89 121

ACMH+	53.71 12	59.26 82	60.28 101	58.06 69	64.17 68	68.46 64	67.51 64	50.93 50	52.46 91	35.83 131	40.83 174	45.12 167	52.32 86	69.88 79	69.00 74	77.59 72	76.21 86

casdiffmvs_mvg		65.26 56	69.48 57	60.33 56	62.99 75	69.34 60	69.80 54	45.27 83	63.38 56	51.11 61	65.12 40	69.75 52	53.51 73	71.74 59	68.86 75	79.33 55	78.19 68
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
UniMVSNet_NR-MVSNet		56.94 108	61.14 91	52.05 116	60.02 93	65.21 102	57.44 128	52.93 40	49.37 107	24.31 179	54.62 86	50.54 128	39.04 149	68.69 88	68.84 76	78.53 63	70.72 115

OMC-MVS		65.16 58	71.35 44	57.94 72	52.95 148	68.82 62	69.00 56	38.28 166	79.89 15	55.20 41	62.76 51	68.31 58	56.14 58	71.30 63	68.70 77	76.06 98	79.67 56

NR-MVSNet		55.35 123	59.46 118	50.56 122	61.33 83	62.97 117	57.91 127	51.80 45	48.62 121	20.59 186	51.99 97	44.73 173	34.10 176	68.58 91	68.64 78	77.66 69	70.67 119

OpenMVS	57.13 9	62.81 65	65.75 70	59.39 62	66.47 56	69.52 59	64.26 96	43.07 128	61.34 62	50.19 65	47.29 123	64.41 68	54.60 64	70.18 77	68.62 79	77.73 68	78.89 61

ET-MVSNet_ETH3D		58.38 94	61.57 89	54.67 95	42.15 198	65.26 99	65.70 82	43.82 107	48.84 114	42.34 98	59.76 62	47.76 139	56.68 53	67.02 124	68.60 80	77.33 76	73.73 105

FC-MVSNet-train		58.40 93	63.15 84	52.85 110	64.29 65	61.84 123	55.98 143	46.47 73	53.06 85	34.96 134	61.95 57	56.37 105	39.49 147	68.67 89	68.36 81	75.92 100	71.81 110

MVS_111021_LR		63.05 64	66.43 65	59.10 64	61.33 83	63.77 114	65.87 81	43.58 115	60.20 64	53.70 53	62.09 56	62.38 75	55.84 60	70.24 76	68.08 82	74.30 114	78.28 67

GeoE		62.43 68	64.79 77	59.68 61	64.15 69	67.17 82	68.80 57	44.42 96	55.65 73	47.38 71	51.54 99	62.51 74	54.04 68	69.99 78	68.07 83	79.28 57	78.57 63

HyFIR lowres test		56.87 109	58.60 127	54.84 93	56.62 125	69.27 61	64.77 92	42.21 134	45.66 142	37.50 126	33.08 192	57.47 99	53.33 78	65.46 144	67.94 84	74.60 111	71.35 112

UniMVSNet (Re)		55.15 128	60.39 100	49.03 136	55.31 130	64.59 106	55.77 144	50.63 52	48.66 120	20.95 185	51.47 100	50.40 129	34.41 175	67.81 108	67.89 85	77.11 80	71.88 109

thisisatest0530		56.68 110	59.68 111	53.19 106	52.97 147	60.96 133	59.41 117	40.51 148	48.26 124	41.06 107	52.67 92	46.30 155	49.78 96	67.66 112	67.83 86	75.39 104	74.07 103

tttt0517		56.53 112	59.59 113	52.95 109	52.66 150	60.99 132	59.21 119	40.51 148	47.89 127	40.40 110	52.50 95	46.04 159	49.78 96	67.75 110	67.83 86	75.15 107	74.17 100

MSDG		58.46 92	58.97 123	57.85 76	66.27 58	66.23 92	67.72 60	42.33 132	53.43 81	43.68 91	43.39 160	45.35 163	49.75 98	68.66 90	67.77 88	77.38 74	67.96 137

DU-MVS		55.41 122	59.59 113	50.54 123	54.60 136	62.97 117	57.44 128	51.80 45	48.62 121	24.31 179	51.99 97	47.00 148	39.04 149	68.11 101	67.75 89	76.03 99	70.72 115

Anonymous202405211			60.60 97		63.44 73	66.71 89	61.00 110	47.23 69	50.62 99		36.85 184	60.63 86	43.03 135	69.17 83	67.72 90	75.41 103	72.54 107

FA-MVS(training)		60.00 79	63.14 85	56.33 86	59.50 97	64.30 110	65.15 89	38.75 163	56.20 71	45.77 80	53.08 89	56.45 102	52.10 89	69.04 86	67.67 91	76.69 83	75.27 95

DCV-MVSNet		59.49 80	64.00 81	54.23 96	61.81 78	64.33 109	61.42 106	43.77 108	52.85 88	38.94 119	55.62 79	62.15 78	43.24 134	69.39 82	67.66 92	76.22 92	75.97 87

UGNet		57.03 105	65.25 73	47.44 155	46.54 184	66.73 86	56.30 138	43.28 124	50.06 100	32.99 138	62.57 53	63.26 72	33.31 178	68.25 96	67.58 93	72.20 146	78.29 66
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
CHOSEN 1792x2688		55.85 117	58.01 131	53.33 103	57.26 119	62.82 119	63.29 102	41.55 139	46.65 134	38.34 120	34.55 190	53.50 113	52.43 85	67.10 122	67.56 94	67.13 168	73.92 104

CANet_DTU		58.88 85	64.68 78	52.12 115	55.77 128	66.75 85	63.92 97	37.04 173	53.32 83	37.45 127	59.81 61	61.81 79	44.43 126	68.25 96	67.47 95	74.12 116	75.33 93

UA-Net		58.50 90	64.68 78	51.30 119	66.97 52	67.13 83	53.68 159	45.65 80	49.51 106	31.58 146	62.91 50	68.47 57	35.85 169	68.20 99	67.28 96	74.03 117	69.24 132

Effi-MVS+-dtu		60.34 77	62.32 87	58.03 71	64.31 64	67.44 79	65.99 79	42.26 133	49.55 104	42.00 101	48.92 113	59.79 89	56.27 56	68.07 103	67.03 97	77.35 75	75.45 92

GBi-Net		55.20 125	60.25 102	49.31 130	52.42 151	61.44 125	57.03 131	44.04 101	49.18 110	30.47 148	48.28 115	58.19 94	38.22 152	68.05 104	66.96 98	73.69 121	69.65 123

test1		55.20 125	60.25 102	49.31 130	52.42 151	61.44 125	57.03 131	44.04 101	49.18 110	30.47 148	48.28 115	58.19 94	38.22 152	68.05 104	66.96 98	73.69 121	69.65 123

FMVSNet1		54.08 133	58.68 125	48.71 140	50.90 167	61.35 128	56.73 135	43.94 106	45.91 140	29.32 158	42.72 168	56.26 106	37.70 159	68.05 104	66.96 98	73.69 121	69.50 127

thisisatest0515		53.85 134	56.84 141	50.37 124	50.25 171	58.17 158	55.99 142	39.90 155	41.88 171	38.16 122	45.91 137	45.30 164	44.58 125	66.15 136	66.89 101	73.36 127	73.57 106

CDS-MVSNet		52.42 141	57.06 140	47.02 157	53.92 143	58.30 156	55.50 147	46.47 73	42.52 168	29.38 157	49.50 108	52.85 118	28.49 188	66.70 127	66.89 101	68.34 163	62.63 173
Khang Truong Giang, Soohwan Song, Sungho Jo: Curvature-guided dynamic scale networks for Multi-view Stereo. ICLR 2022
MVSTER		57.19 104	61.11 92	52.62 112	50.82 168	58.79 150	61.55 104	37.86 169	48.81 116	41.31 104	57.43 73	52.10 120	48.60 104	68.19 100	66.75 103	75.56 102	75.68 91

MS-PatchMatch		58.19 99	60.20 104	55.85 89	65.17 61	64.16 111	64.82 91	41.48 140	50.95 96	42.17 100	45.38 144	56.42 103	48.08 108	68.30 95	66.70 104	73.39 125	69.46 130

DI_MVS_plusplus_trai		61.88 70	65.17 74	58.06 69	60.05 91	65.26 99	66.03 78	44.22 97	55.75 72	46.73 74	54.64 85	68.12 61	54.13 67	69.13 84	66.66 105	77.18 77	76.61 79

TSAR-MVS + COLMAP		62.65 67	69.90 53	54.19 97	46.31 185	66.73 86	65.49 86	41.36 141	76.57 24	46.31 77	76.80 17	56.68 100	53.27 80	69.50 81	66.65 106	72.40 143	76.36 85

Anonymous20231211		57.71 102	60.79 94	54.13 98	61.68 81	65.81 95	60.81 111	43.70 112	51.97 94	39.67 114	34.82 189	63.59 70	43.31 132	68.55 93	66.63 107	75.59 101	74.13 101

v1144		58.88 85	60.16 105	57.39 78	58.03 104	67.26 80	67.14 67	44.46 94	45.17 144	44.33 89	47.81 120	49.92 132	53.20 81	67.77 109	66.62 108	77.15 78	76.58 80

v10		59.17 84	60.60 97	57.50 77	57.95 105	66.73 86	67.09 69	44.11 98	46.85 132	45.42 83	48.18 119	51.07 124	53.63 70	67.84 107	66.59 109	76.79 81	76.92 76

v1192		58.51 89	59.66 112	57.17 79	57.82 106	67.72 73	66.21 77	44.83 89	44.15 152	43.49 92	46.68 125	47.94 136	53.55 72	67.39 116	66.51 110	77.13 79	77.20 74

LS3D		60.20 78	61.70 88	58.45 67	64.18 67	67.77 72	67.19 65	48.84 65	61.67 61	41.27 105	45.89 138	51.81 122	54.18 66	68.78 87	66.50 111	75.03 109	69.48 128

Fast-Effi-MVS+-dtu		56.30 114	59.29 120	52.82 111	58.64 101	64.89 103	65.56 85	32.89 195	45.80 141	35.04 133	45.89 138	54.14 112	49.41 99	67.16 120	66.45 112	75.37 105	70.69 117

MVS_Test		62.40 69	66.23 67	57.94 72	59.77 96	64.77 105	66.50 74	41.76 137	57.26 70	49.33 66	62.68 52	67.47 64	53.50 75	68.57 92	66.25 113	76.77 82	76.58 80

v7n		55.67 119	57.46 138	53.59 102	56.06 126	65.29 98	61.06 109	43.26 125	40.17 182	37.99 123	40.79 175	45.27 166	47.09 113	67.67 111	66.21 114	76.08 95	76.82 77

FMVSNet2		55.04 129	59.95 110	49.31 130	52.42 151	61.44 125	57.03 131	44.08 100	49.55 104	30.40 151	46.89 124	58.84 92	38.22 152	67.07 123	66.21 114	73.69 121	69.65 123

casdiffmvs		64.09 60	68.13 61	59.37 63	61.81 78	68.32 66	68.48 59	44.45 95	61.95 60	49.12 68	63.04 49	69.67 54	53.83 69	70.46 72	66.06 116	78.55 62	77.43 71
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
v2v482		58.69 88	60.12 108	57.03 80	57.16 122	66.05 93	67.17 66	43.52 117	46.33 136	45.19 85	49.46 109	51.02 125	52.51 84	67.30 117	66.03 117	76.61 84	74.62 97

Baseline_NR-MVSNet		53.50 135	57.89 132	48.37 146	54.60 136	59.25 147	56.10 139	51.84 44	49.32 108	17.92 194	45.38 144	47.68 140	36.93 164	68.11 101	65.95 118	72.84 133	69.57 126

V42		56.97 107	60.14 106	53.28 104	48.16 177	62.78 120	66.30 76	37.93 168	47.44 129	42.68 96	48.19 118	52.59 119	51.90 90	67.46 115	65.94 119	72.72 136	76.55 82

v144192		58.23 98	59.40 119	56.87 82	57.56 107	66.89 84	65.70 82	45.01 87	44.06 153	42.88 94	46.61 127	48.09 135	53.49 76	66.94 125	65.90 120	76.61 84	77.29 72

v1240		57.55 103	58.63 126	56.29 87	57.30 118	66.48 91	63.77 98	44.56 93	42.77 166	42.48 97	45.64 141	46.28 156	53.46 77	66.32 132	65.80 121	76.16 93	77.13 75

v1921920		57.89 101	59.02 122	56.58 85	57.55 108	66.66 90	64.72 93	44.70 91	43.55 156	42.73 95	46.17 135	46.93 149	53.51 73	66.78 126	65.75 122	76.29 89	77.28 73

v8		58.88 85	60.57 99	56.92 81	57.35 115	65.69 96	66.69 73	42.64 130	47.89 127	45.77 80	49.04 110	52.98 117	52.77 82	67.51 114	65.57 123	76.26 91	75.30 94

diffmvs		61.64 71	66.55 64	55.90 88	56.63 124	63.71 115	67.13 68	41.27 142	59.49 65	46.70 75	63.93 47	68.01 62	50.46 95	67.30 117	65.51 124	73.24 131	77.87 70
Fangjinhua Wang, Qingshan Xu, Yew-Soon Ong, Marc Pollefeys: Lightweight and Accurate Multi-View Stereo With Confidence-Aware Diffusion Model. IEEE T-PAMI 2025
GA-MVS		55.67 119	58.33 128	52.58 113	55.23 133	63.09 116	61.08 108	40.15 154	42.95 161	37.02 129	52.61 93	47.68 140	47.51 111	65.92 138	65.35 125	74.49 113	70.68 118

IterMVS-LS		58.30 96	61.39 90	54.71 94	59.92 94	58.40 154	59.42 116	43.64 113	48.71 118	40.25 112	57.53 71	58.55 93	52.15 88	65.42 145	65.34 126	72.85 132	75.77 88
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
PLC	52.09 14	59.21 83	62.47 86	55.41 92	53.24 146	64.84 104	64.47 95	40.41 152	65.92 51	44.53 88	46.19 134	55.69 108	55.33 62	68.24 98	65.30 127	74.50 112	71.09 113
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
CNLPA		62.78 66	66.31 66	58.65 66	58.47 102	68.41 65	65.98 80	41.22 143	78.02 22	56.04 36	46.65 126	59.50 90	57.50 45	69.67 80	65.27 128	72.70 138	76.67 78

TransMVSNet (Re)		51.92 148	55.38 147	47.88 152	60.95 86	59.90 141	53.95 156	45.14 86	39.47 185	24.85 176	43.87 155	46.51 154	29.15 185	67.55 113	65.23 129	73.26 130	65.16 161

PVSNet_BlendedMVS		61.63 72	64.82 75	57.91 74	57.21 120	67.55 77	63.47 100	46.08 75	54.72 75	52.46 57	58.59 66	60.73 83	51.82 92	70.46 72	65.20 130	76.44 87	76.50 83

PVSNet_Blended		61.63 72	64.82 75	57.91 74	57.21 120	67.55 77	63.47 100	46.08 75	54.72 75	52.46 57	58.59 66	60.73 83	51.82 92	70.46 72	65.20 130	76.44 87	76.50 83

FMVSNet3		54.78 130	59.58 115	49.17 133	52.37 154	61.31 129	56.72 136	44.04 101	49.18 110	30.47 148	48.28 115	58.19 94	38.09 155	65.48 143	65.20 130	73.31 128	69.45 131

UniMVSNet_ETH3D		52.62 139	55.98 143	48.70 141	51.04 165	60.71 135	56.87 134	46.74 72	42.52 168	26.96 169	42.50 170	45.95 160	37.87 156	66.22 134	65.15 133	72.74 135	68.78 135

TAPA-MVS	54.74 10	60.85 75	66.61 63	54.12 99	47.38 181	65.33 97	65.35 87	36.51 175	75.16 30	48.82 70	54.70 84	63.51 71	53.31 79	68.36 94	64.97 134	73.37 126	74.27 99
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
WR-MVS		48.78 168	55.06 150	41.45 182	55.50 129	60.40 136	43.77 197	49.99 56	41.92 170	8.10 213	45.24 147	45.56 161	17.47 202	61.57 161	64.60 135	73.85 118	66.14 153

tfpn200view9		52.53 140	55.51 145	49.06 135	57.31 116	60.24 137	55.42 149	43.77 108	42.85 164	27.81 163	43.00 166	45.06 169	37.32 161	66.38 129	64.54 136	72.71 137	66.54 146

thres400		52.38 143	55.51 145	48.74 139	57.49 111	60.10 140	55.45 148	43.54 116	42.90 163	26.72 171	43.34 162	45.03 171	36.61 166	66.20 135	64.53 137	72.66 139	66.43 147

gg-mvs-nofinetune		49.07 166	52.56 166	45.00 166	61.99 76	59.78 142	53.55 161	41.63 138	31.62 207	12.08 202	29.56 201	53.28 116	29.57 184	66.27 133	64.49 138	71.19 154	62.92 170

baseline2		55.89 115	57.82 133	53.64 100	57.36 114	61.09 131	59.75 115	40.45 150	47.38 130	41.26 106	51.23 101	46.90 150	48.11 107	65.63 142	64.38 139	74.90 110	68.16 136

baseline1		54.48 132	58.69 124	49.57 128	60.63 88	58.29 157	55.70 145	44.95 88	49.20 109	29.62 155	54.77 83	54.75 110	35.29 170	67.15 121	64.08 140	71.21 153	62.58 174

PEN-MVS		49.21 164	54.32 154	43.24 176	54.33 139	59.26 146	47.04 183	51.37 49	41.67 172	9.97 208	46.22 133	41.80 184	22.97 198	60.52 164	64.03 141	73.73 120	66.75 145

pm-mvs1		51.02 152	55.55 144	45.73 161	54.16 140	58.52 152	50.92 166	42.56 131	40.32 180	25.67 174	43.66 157	50.34 130	30.06 183	65.85 139	63.97 142	70.99 155	66.21 150

thres200		52.39 142	55.37 148	48.90 137	57.39 113	60.18 138	55.60 146	43.73 110	42.93 162	27.41 165	43.35 161	45.09 168	36.61 166	66.36 130	63.92 143	72.66 139	65.78 156

pmmvs4		54.66 131	56.07 142	53.00 108	54.63 135	57.08 165	60.43 113	44.10 99	51.69 95	40.55 109	46.55 130	44.79 172	45.95 119	62.54 155	63.66 144	72.36 144	66.20 151

thres100view900		52.04 146	54.81 152	48.80 138	57.31 116	59.33 145	55.30 150	42.92 129	42.85 164	27.81 163	43.00 166	45.06 169	36.99 163	64.74 148	63.51 145	72.47 142	65.21 160

tfpnnormal		50.16 158	52.19 170	47.78 154	56.86 123	58.37 155	54.15 155	44.01 104	38.35 193	25.94 173	36.10 185	37.89 199	34.50 174	65.93 137	63.42 146	71.26 152	65.28 159

Vis-MVSNet (Re-imp)		50.37 156	57.73 136	41.80 181	57.53 109	54.35 171	45.70 189	45.24 84	49.80 102	13.43 200	58.23 69	56.42 103	20.11 201	62.96 153	63.36 147	68.76 162	58.96 186

thres600view7		51.91 149	55.14 149	48.14 148	57.43 112	60.18 138	54.60 154	43.73 110	42.61 167	25.20 175	43.10 165	44.47 176	35.19 171	66.36 130	63.28 148	72.66 139	66.01 154

pmmvs6		48.35 170	51.64 172	44.51 169	51.92 157	57.94 161	49.44 172	42.17 135	34.45 200	24.62 178	28.87 203	46.90 150	29.07 187	64.60 149	63.08 149	69.83 159	65.68 157

COLMAP_ROB	46.52 15	51.99 147	54.86 151	48.63 142	49.13 175	61.73 124	60.53 112	36.57 174	53.14 84	32.95 139	37.10 182	38.68 197	40.49 144	65.72 140	63.08 149	72.11 147	64.60 164
Johannes L. SchÃ¶nberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
gm-plane-assit		44.74 186	45.95 194	43.33 174	60.88 87	46.79 199	36.97 208	32.24 198	24.15 213	11.79 203	29.26 202	32.97 208	46.64 114	65.09 147	62.95 151	71.45 151	60.42 181

DTE-MVSNet		48.03 174	53.28 161	41.91 180	54.64 134	57.50 163	44.63 196	51.66 48	41.02 176	7.97 214	46.26 132	40.90 187	20.24 200	60.45 165	62.89 152	72.33 145	63.97 166

pmmvs-eth3d		51.33 150	52.25 169	50.26 125	50.82 168	54.65 170	56.03 141	43.45 122	43.51 157	37.20 128	39.20 178	39.04 196	42.28 137	61.85 160	62.78 153	71.78 149	64.72 163

LTVRE_ROB	44.17 16	47.06 180	50.15 183	43.44 173	51.39 160	58.42 153	42.90 199	43.51 118	22.27 215	14.85 198	41.94 173	34.57 205	45.43 120	62.28 158	62.77 154	62.56 185	68.83 134
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
EPNet_dtu		52.05 145	58.26 129	44.81 167	54.10 141	50.09 186	52.01 164	40.82 146	53.03 86	27.41 165	54.90 81	57.96 98	26.72 190	62.97 152	62.70 155	67.78 166	66.19 152
Wanjuan Su, Wenbing Tao: Efficient Edge-Preserving Multi-View Stereo Network for Depth Estimation. AAAI 2023
v148		55.58 121	57.61 137	53.20 105	54.59 138	61.86 122	61.18 107	38.70 164	44.30 151	42.25 99	47.53 121	50.24 131	48.73 102	65.15 146	62.61 156	73.79 119	71.61 111

TDRefinement		49.31 161	52.44 167	45.67 163	30.44 211	59.42 144	59.24 118	39.78 156	48.76 117	31.20 147	35.73 186	29.90 211	42.81 136	64.24 150	62.59 157	70.55 156	66.43 147

CP-MVSNet		48.37 169	53.53 158	42.34 178	51.35 161	58.01 160	46.56 184	50.54 53	41.62 173	10.61 204	46.53 131	40.68 190	23.18 196	58.71 174	61.83 158	71.81 148	67.36 143

PS-CasMVS		48.18 171	53.25 162	42.27 179	51.26 162	57.94 161	46.51 185	50.52 54	41.30 174	10.56 205	45.35 146	40.34 192	23.04 197	58.66 175	61.79 159	71.74 150	67.38 141

IterMVS-SCA-FT		52.18 144	57.75 135	45.68 162	51.01 166	62.06 121	55.10 152	34.75 181	44.85 145	32.86 140	51.13 103	51.22 123	48.74 101	62.47 156	61.51 160	51.61 207	71.02 114

WR-MVS_H		47.65 175	53.67 157	40.63 185	51.45 159	59.74 143	44.71 195	49.37 58	40.69 178	7.61 215	46.04 136	44.34 178	17.32 203	57.79 180	61.18 161	73.30 129	65.86 155

USDC		51.11 151	53.71 156	48.08 150	44.76 190	55.99 168	53.01 163	40.90 144	52.49 90	36.14 130	44.67 149	33.66 207	43.27 133	63.23 151	61.10 162	70.39 158	64.82 162

SixPastTwentyTwo		47.55 177	50.25 182	44.41 170	47.30 182	54.31 172	47.81 178	40.36 153	33.76 201	19.93 189	43.75 156	32.77 209	42.07 138	59.82 167	60.94 163	68.98 160	66.37 149

IterMVS		53.45 136	57.12 139	49.17 133	49.23 174	60.93 134	59.05 120	34.63 183	44.53 147	33.22 136	51.09 104	51.01 126	48.38 105	62.43 157	60.79 164	70.54 157	69.05 133
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Marc Pollefeys: IterMVS: Iterative Probability Estimation for Efficient Multi-View Stereo.
baseline		55.19 127	60.88 93	48.55 143	49.87 172	58.10 159	58.70 121	34.75 181	52.82 89	39.48 118	60.18 60	60.86 82	45.41 121	61.05 162	60.74 165	63.10 181	72.41 108

CR-MVSNet		50.47 154	52.61 165	47.98 151	49.03 176	52.94 175	48.27 175	38.86 160	44.41 148	39.59 115	44.34 151	44.65 175	46.63 115	58.97 171	60.31 166	65.48 173	62.66 171

PatchT		48.08 172	51.03 177	44.64 168	42.96 195	50.12 185	40.36 204	35.09 179	43.17 159	39.59 115	42.00 172	39.96 193	46.63 115	58.97 171	60.31 166	63.21 180	62.66 171

CMPMVS	37.70 17	49.24 163	52.71 164	45.19 164	45.97 187	51.23 182	47.44 181	29.31 200	43.04 160	44.69 86	34.45 191	48.35 134	43.64 128	62.59 154	59.82 168	60.08 189	69.48 128
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
PatchMatch-RL		50.11 159	51.56 173	48.43 144	46.23 186	51.94 179	50.21 169	38.62 165	46.62 135	37.51 125	42.43 171	39.38 194	52.24 87	60.98 163	59.56 169	65.76 172	60.01 184

TinyColmap		47.08 178	47.56 192	46.52 158	42.35 197	53.44 174	51.77 165	40.70 147	43.44 158	31.92 144	29.78 200	23.72 217	45.04 124	61.99 159	59.54 170	67.35 167	61.03 178

PMMVS		49.20 165	54.28 155	43.28 175	34.13 206	45.70 201	48.98 173	26.09 208	46.31 137	34.92 135	55.22 80	53.47 114	47.48 112	59.43 168	59.04 171	68.05 165	60.77 179

pmmvs5		47.07 179	51.02 178	42.46 177	45.18 189	51.47 181	48.23 177	33.09 194	38.17 194	28.62 161	46.60 128	43.48 180	30.74 181	58.28 177	58.63 172	68.92 161	60.48 180

CostFormer		56.57 111	59.13 121	53.60 101	57.52 110	61.12 130	66.94 71	35.95 177	53.44 80	44.68 87	55.87 78	54.44 111	48.21 106	60.37 166	58.33 173	68.27 164	70.33 120

ambc			45.54 198		50.66 170	52.63 178	40.99 203		38.36 192	24.67 177	22.62 210	13.94 220	29.14 186	65.71 141	58.06 174	58.60 193	67.43 139

TAMVS		44.02 189	49.18 186	37.99 193	47.03 183	45.97 200	45.04 192	28.47 203	39.11 188	20.23 188	43.22 164	48.52 133	28.49 188	58.15 178	57.95 175	58.71 191	51.36 197

SCA		50.99 153	53.22 163	48.40 145	51.07 164	56.78 166	50.25 168	39.05 157	48.31 123	41.38 103	49.54 107	46.70 153	46.00 118	58.31 176	56.28 176	62.65 183	56.60 191

dps		50.42 155	51.20 176	49.51 129	55.88 127	56.07 167	53.73 157	38.89 159	43.66 154	40.36 111	45.66 140	37.63 201	45.23 122	59.05 169	56.18 177	62.94 182	60.16 182

MDA-MVSNet-bldmvs		41.36 194	43.15 204	39.27 189	28.74 213	52.68 177	44.95 194	40.84 145	32.89 203	18.13 193	31.61 195	22.09 218	38.97 151	50.45 206	56.11 178	64.01 178	56.23 192

MIMVSNet		43.79 190	48.53 188	38.27 191	41.46 199	48.97 189	50.81 167	32.88 196	44.55 146	22.07 182	32.05 193	47.15 146	24.76 193	58.73 173	56.09 179	57.63 196	52.14 195

test-mter		45.30 185	50.37 179	39.38 188	33.65 208	46.99 196	47.59 179	18.59 214	38.75 189	28.00 162	43.28 163	46.82 152	41.50 141	57.28 182	55.78 180	66.93 171	63.70 168

CVMVSNet		46.38 183	52.01 171	39.81 187	42.40 196	50.26 184	46.15 186	37.68 170	40.03 183	15.09 197	46.56 129	47.56 142	33.72 177	56.50 188	55.65 181	63.80 179	67.53 138

MDTV_nov1_ep13_2view		47.62 176	49.72 185	45.18 165	48.05 178	53.70 173	54.90 153	33.80 189	39.90 184	29.79 154	38.85 179	41.89 183	39.17 148	58.99 170	55.55 182	65.34 175	59.17 185

test-LLR		49.28 162	50.29 180	48.10 149	55.26 131	47.16 194	49.52 170	43.48 120	39.22 186	31.98 142	43.65 158	47.93 137	41.29 142	56.80 184	55.36 183	67.08 169	61.94 175

TESTMET0.1,1		46.09 184	50.29 180	41.18 183	36.91 204	47.16 194	49.52 170	20.32 213	39.22 186	31.98 142	43.65 158	47.93 137	41.29 142	56.80 184	55.36 183	67.08 169	61.94 175

MDTV_nov1_ep13		50.32 157	52.43 168	47.86 153	49.87 172	54.70 169	58.10 125	34.29 185	45.59 143	37.71 124	47.44 122	47.42 144	41.86 139	58.07 179	55.21 185	65.34 175	58.56 187

FMVSNet5		40.96 195	45.81 196	35.29 200	34.30 205	44.55 204	47.28 182	28.84 202	40.76 177	21.62 183	29.85 199	42.44 181	24.77 192	57.53 181	55.00 186	54.93 199	50.56 200

FC-MVSNet-test		39.65 201	48.35 189	29.49 205	44.43 191	39.28 211	30.23 214	40.44 151	43.59 155	3.12 221	53.00 90	42.03 182	10.02 217	55.09 195	54.77 187	48.66 209	50.71 199

test20.03		40.38 200	44.20 200	35.92 198	53.73 144	49.05 187	38.54 206	43.49 119	32.55 204	9.54 209	27.88 204	39.12 195	12.24 209	56.28 189	54.69 188	57.96 195	49.83 205

Anonymous20231206		42.28 192	45.89 195	38.07 192	51.96 156	48.98 188	43.66 198	38.81 162	38.74 190	14.32 199	26.74 205	40.90 187	20.94 199	56.64 187	54.67 189	58.71 191	54.59 193

CHOSEN 280x420		40.80 196	45.05 199	35.84 199	32.95 209	29.57 214	44.98 193	23.71 211	37.54 196	18.42 192	31.36 196	47.07 147	46.41 117	56.71 186	54.65 190	48.55 210	58.47 188

test0.0.03 1		43.15 191	46.95 193	38.72 190	55.26 131	50.56 183	42.48 200	43.48 120	38.16 195	15.11 196	35.07 188	44.69 174	16.47 204	55.95 192	54.34 191	59.54 190	49.87 204

tpm cat1		53.30 137	53.41 159	53.17 107	58.16 103	59.15 148	63.73 99	38.27 167	50.73 98	46.98 73	45.57 142	44.00 179	49.20 100	55.90 193	54.02 192	62.65 183	64.50 165

RPMNet		46.41 181	48.72 187	43.72 171	47.77 180	52.94 175	46.02 188	33.92 187	44.41 148	31.82 145	36.89 183	37.42 202	37.41 160	53.88 199	54.02 192	65.37 174	61.47 177

MIMVSNet1		35.51 205	41.41 205	28.63 206	27.53 215	43.36 205	38.09 207	33.82 188	32.01 205	6.77 216	21.63 211	35.43 204	11.97 211	55.05 196	53.99 194	53.59 204	48.36 207

PatchmatchNet		49.92 160	51.29 174	48.32 147	51.83 158	51.86 180	53.38 162	37.63 171	47.90 126	40.83 108	48.54 114	45.30 164	45.19 123	56.86 183	53.99 194	61.08 188	54.57 194
Fangjinhua Wang, Silvano Galliani, Christoph Vogel, Pablo Speciale, Marc Pollefeys: PatchmatchNet: Learned Multi-View Patchmatch Stereo.
PM-MVS		44.55 188	48.13 190	40.37 186	32.85 210	46.82 198	46.11 187	29.28 201	40.48 179	29.99 153	39.98 177	34.39 206	41.80 140	56.08 191	53.88 196	62.19 186	65.31 158

testgi		38.71 202	43.64 202	32.95 202	52.30 155	48.63 190	35.59 211	35.05 180	31.58 208	9.03 212	30.29 197	40.75 189	11.19 215	55.30 194	53.47 197	54.53 202	45.48 208

RPSCF		46.41 181	54.42 153	37.06 195	25.70 218	45.14 202	45.39 191	20.81 212	62.79 58	35.10 132	44.92 148	55.60 109	43.56 129	56.12 190	52.45 198	51.80 206	63.91 167

pmmvs3		35.10 206	38.47 208	31.17 204	26.37 217	40.47 207	34.51 212	18.09 215	24.75 212	16.88 195	23.05 209	26.69 213	32.69 179	50.73 205	51.60 199	58.46 194	51.98 196

GG-mvs-BLEND		36.62 204	53.39 160	17.06 212	0.01 224	58.61 151	48.63 174	0.01 221	47.13 131	0.02 225	43.98 153	60.64 85	0.03 220	54.92 197	51.47 200	53.64 203	56.99 190

tpm		48.82 167	51.27 175	45.96 160	54.10 141	47.35 193	56.05 140	30.23 199	46.70 133	43.21 93	52.54 94	47.55 143	37.28 162	54.11 198	50.50 201	54.90 200	60.12 183

EU-MVSNet		40.63 198	45.65 197	34.78 201	39.11 202	46.94 197	40.02 205	34.03 186	33.50 202	10.37 206	35.57 187	37.80 200	23.65 195	51.90 201	50.21 202	61.49 187	63.62 169

tpmrst		48.08 172	49.88 184	45.98 159	52.71 149	48.11 191	53.62 160	33.70 190	48.70 119	39.74 113	48.96 112	46.23 157	40.29 146	50.14 207	49.28 203	55.80 197	57.71 189

Gipuma		25.87 210	26.91 213	24.66 209	28.98 212	20.17 217	20.46 216	34.62 184	29.55 209	9.10 210	4.91 220	5.31 224	15.76 206	49.37 210	49.10 204	39.03 213	29.95 213
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
EPMVS		44.66 187	47.86 191	40.92 184	47.97 179	44.70 203	47.58 180	33.27 192	48.11 125	29.58 156	49.65 106	44.38 177	34.65 172	51.71 202	47.90 205	52.49 205	48.57 206

PMVS	27.84 18	33.81 207	35.28 211	32.09 203	34.13 206	24.81 216	32.51 213	26.48 207	26.41 211	19.37 190	23.76 208	24.02 216	25.18 191	50.78 203	47.24 206	54.89 201	49.95 203
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
MVS-HIRNet		42.24 193	41.15 206	43.51 172	44.06 194	40.74 206	35.77 210	35.35 178	35.38 199	38.34 120	25.63 207	38.55 198	43.48 130	50.77 204	47.03 207	64.07 177	49.98 202

FPMVS		38.36 203	40.41 207	35.97 197	38.92 203	39.85 209	45.50 190	25.79 209	41.13 175	18.70 191	30.10 198	24.56 215	31.86 180	49.42 209	46.80 208	55.04 198	51.03 198

pmnet_mix02		40.48 199	43.80 201	36.61 196	45.79 188	40.45 208	42.12 201	33.18 193	40.30 181	24.11 181	38.76 180	37.11 203	24.30 194	52.97 200	46.66 209	50.17 208	50.33 201

ADS-MVSNet		40.67 197	43.38 203	37.50 194	44.36 192	39.79 210	42.09 202	32.67 197	44.34 150	28.87 160	40.76 176	40.37 191	30.22 182	48.34 211	45.87 210	46.81 211	44.21 210

N_pmnet		32.67 209	36.85 210	27.79 208	40.55 200	32.13 213	35.80 209	26.79 206	37.24 197	9.10 210	32.02 194	30.94 210	16.30 205	47.22 212	41.21 211	38.21 214	37.21 211

new-patchmatchnet		33.24 208	37.20 209	28.62 207	44.32 193	38.26 212	29.68 215	36.05 176	31.97 206	6.33 217	26.59 206	27.33 212	11.12 216	50.08 208	41.05 212	44.23 212	45.15 209

new_pmnet		23.19 211	28.17 212	17.37 210	17.03 219	24.92 215	19.66 217	16.16 217	27.05 210	4.42 218	20.77 212	19.20 219	12.19 210	37.71 213	36.38 213	34.77 215	31.17 212

MVE	12.28 19	13.53 215	15.72 215	10.96 215	7.39 221	15.71 219	6.05 222	23.73 210	10.29 221	3.01 222	5.77 219	3.41 225	11.91 212	20.11 215	29.79 214	13.67 220	24.98 214
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
PMMVS2		15.84 212	19.68 214	11.35 214	15.74 220	16.95 218	13.31 218	17.64 216	16.08 217	0.36 224	13.12 214	11.47 221	1.69 219	28.82 214	27.24 215	19.38 219	24.09 215

E-PMN		15.09 213	13.19 217	17.30 211	27.80 214	12.62 220	7.81 221	27.54 204	14.62 219	3.19 219	6.89 217	2.52 227	15.09 207	15.93 217	20.22 216	22.38 216	19.53 216

EMVS		14.49 214	12.45 218	16.87 213	27.02 216	12.56 221	8.13 220	27.19 205	15.05 218	3.14 220	6.69 218	2.67 226	15.08 208	14.60 219	18.05 217	20.67 217	17.56 218

tmp_tt				5.40 217	3.97 222	2.35 224	3.26 224	0.44 220	17.56 216	12.09 201	11.48 216	7.14 222	1.98 218	15.68 218	15.49 218	10.69 221

test_method		12.44 216	14.66 216	9.85 216	1.30 223	3.32 223	13.00 219	3.21 218	22.42 214	10.22 207	14.13 213	25.64 214	11.43 214	19.75 216	11.61 219	19.96 218	5.79 219

testmvs		0.01 217	0.02 219	0.00 218	0.00 225	0.00 225	0.01 226	0.00 222	0.01 222	0.00 226	0.03 222	0.00 228	0.01 221	0.01 221	0.01 220	0.00 223	0.06 221

test123		0.01 217	0.02 219	0.00 218	0.00 225	0.00 225	0.00 227	0.00 222	0.01 222	0.00 226	0.04 221	0.00 228	0.01 221	0.00 222	0.01 220	0.00 223	0.07 220

uanet_test		0.00 219	0.00 221	0.00 218	0.00 225	0.00 225	0.00 227	0.00 222	0.00 224	0.00 226	0.00 223	0.00 228	0.00 223	0.00 222	0.00 222	0.00 223	0.00 222

sosnet-low-res		0.00 219	0.00 221	0.00 218	0.00 225	0.00 225	0.00 227	0.00 222	0.00 224	0.00 226	0.00 223	0.00 228	0.00 223	0.00 222	0.00 222	0.00 223	0.00 222

sosnet		0.00 219	0.00 221	0.00 218	0.00 225	0.00 225	0.00 227	0.00 222	0.00 224	0.00 226	0.00 223	0.00 228	0.00 223	0.00 222	0.00 222	0.00 223	0.00 222

RE-MVS-def										33.01 137

9.14												81.81 13

SR-MVS					71.46 33			54.67 28				81.54 14

our_test_3					51.15 163	57.31 164	55.12 151

MTAPA										65.14 4		80.20 20

MTMP										62.63 16		78.04 26

Patchmatch-RL test							1.04 225

XVS					70.49 37	76.96 25	74.36 44			54.48 47		74.47 37				82.24 24

X-MVStestdata					70.49 37	76.96 25	74.36 44			54.48 47		74.47 37				82.24 24

mPP-MVS					71.67 32							74.36 40

NP-MVS									72.00 39

Patchmtry						47.61 192	48.27 175	38.86 160		39.59 115

DeepMVS_CX						6.95 222	5.98 223	2.25 219	11.73 220	2.07 223	11.85 215	5.43 223	11.75 213	11.40 220		8.10 222	18.38 217

High-resolution multi-view benchmark