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.

PatchmatchNet		492.52 1	551.22 1	316.40 2	400.20 3	357.90 2	1277.10 3	115.80 3	1034.60 4	315.90 2	874.00 2	155.10 6	402.20 3	629.50 2	158.80 2	189.10 2

ACMH+	729.97 1	661.38 2	754.00 2	383.51 3	368.96 2	376.38 3	2483.74 6	82.42 2	1089.25 5	333.39 3	1219.68 6	66.53 2	573.60 5	957.86 6	184.14 3	200.56 3

Gipuma		689.75 3	767.00 3	458.00 4	613.00 9	549.00 6	1943.00 4	187.00 10	1230.00 6	518.00 6	1092.00 5	203.00 12	564.00 4	892.00 5	225.00 4	261.00 4
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PVSNet	1207.94 4	829.56 4	937.18 4	506.69 6	645.55 10	585.72 7	2477.98 5	130.52 4	1578.29 7	509.50 5	1460.01 7	194.90 9	637.19 6	1170.78 9	267.07 7	297.17 6

PMVS	863.65 2	957.08 5	1054.11 7	666.00 10	401.00 4	495.00 4	3790.00 9	279.00 13	1607.00 9	521.00 7	1745.00 9	90.00 3	1105.00 14	639.00 3	415.00 12	398.00 11
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
ACMH	993.80 3	967.92 6	1106.11 9	553.33 7	494.00 5	534.00 5	3581.00 8	136.00 5	1585.00 8	478.00 4	1871.00 10	90.00 3	839.00 7	1460.00 11	260.00 6	287.00 5
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
PVSNet_0	1661.03 7	968.54 7	1066.82 8	673.70 11	706.94 12	695.69 9	2669.02 7	153.52 7	2108.17 13	605.54 9	1679.14 8	213.26 14	886.75 8	1236.76 10	229.01 5	438.67 12

TinyColmap		1000.00 8	1000.00 5	1000.00 13	1000.00 14	1000.00 14	1000.00 1	1000.00 34	1000.00 2	1000.00 15	1000.00 3	1000.00 36	1000.00 11	1000.00 7	1000.00 26	1000.00 26

USDC		1000.00 8	1000.00 5	1000.00 13	1000.00 14	1000.00 14	1000.00 1	1000.00 34	1000.00 2	1000.00 15	1000.00 3	1000.00 36	1000.00 11	1000.00 7	1000.00 26	1000.00 26

ACMP	1246.57 5	1086.24 10	1235.17 10	639.45 8	596.53 7	639.91 8	3904.24 10	152.04 6	1851.54 10	566.58 8	1997.95 11	138.50 5	938.30 9	1605.46 13	303.70 9	340.14 8
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
ACMM	1311.08 6	1165.06 11	1332.72 11	662.07 9	584.90 6	715.99 10	4373.59 12	153.99 8	1880.27 11	709.02 11	2316.97 12	188.63 8	939.89 10	1484.30 12	302.78 8	330.34 7
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
MG-MVS		1256.50 12	1432.00 12	730.00 12	610.00 8	760.00 11	4720.00 13	169.00 9	1920.00 12	723.00 12	2510.00 15	201.00 11	1050.00 13	1645.00 14	390.00 11	380.00 9

FPMVS		1407.00 13	1495.59 13	1141.23 16	648.01 11	893.85 13	4124.00 11	503.81 23	3219.04 17	742.72 13	2498.84 14	157.98 7	1727.83 18	832.93 4	732.98 17	802.00 16

DeepC-MVS_fast	2190.19 10	1535.59 14	1886.82 15	481.89 5	847.73 13	830.16 12	7943.84 21	189.23 11	2398.83 14	691.73 10	2341.13 13	209.16 13	230.25 2	1999.25 15	360.46 10	385.25 10
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
COLMAP_ROB	2144.59 9	1658.33 15	1869.33 14	1025.33 15	1030.00 16	1023.00 16	5311.00 15	359.00 14	3044.00 16	937.00 14	2892.00 16	332.00 17	1440.00 15	2312.00 16	607.00 13	613.00 14
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
F-COLMAP		1878.17 16	2113.11 16	1173.33 18	1231.00 17	1174.00 17	6297.00 16	392.00 16	3356.00 19	1087.00 17	3082.00 17	361.00 19	1670.00 17	2523.00 18	689.00 14	676.00 15

OpenMVS_ROB	1936.51 8	1959.51 17	2224.88 18	1163.39 17	1253.91 18	1434.22 20	7065.71 17	272.76 12	2472.29 15	1647.54 27	3538.72 19	200.37 10	1496.48 16	2575.17 19	997.45 25	559.48 13

CMPMVS	2623.62 13	1983.08 18	2156.14 17	1463.89 23	1414.80 20	1517.00 21	5239.20 14	867.40 31	3310.30 18	1128.70 19	3489.90 18	366.75 20	2030.20 23	2427.60 17	1160.60 29	844.46 20
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
PLC	2858.82 14	2192.92 19	2474.00 19	1349.67 20	1407.00 19	1356.00 19	7164.00 18	475.00 21	3970.00 21	1237.00 22	3765.00 22	451.00 21	1891.00 20	3006.00 21	791.00 18	802.00 16
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
OpenMVS	2352.88 11	2263.08 20	2530.89 20	1459.67 22	1767.00 24	1591.00 24	7932.00 20	399.00 17	3643.00 20	1176.00 20	3569.00 20	297.00 16	1927.00 22	3302.00 22	693.00 15	861.00 21

PCF-MVS	4335.85 18	2272.23 21	2597.03 21	1297.84 19	1701.20 21	1332.37 18	8105.11 22	451.45 18	4031.84 22	1099.10 18	3966.50 23	467.45 22	1734.05 19	2701.68 20	848.94 21	827.11 19
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
3Dnovator	2482.06 12	2531.17 22	2839.56 22	1606.00 24	1732.00 22	1557.00 23	8375.00 24	483.00 22	4370.00 23	1371.00 24	4200.00 24	350.00 18	2207.00 24	3877.00 26	798.00 19	1054.00 28

AdaColmap		2570.77 23	2948.61 23	1437.24 21	1873.64 25	1526.60 22	7819.77 19	373.97 15	4970.45 26	1479.29 25	4648.46 26	525.80 25	1904.97 21	4150.42 27	695.69 16	880.14 22

DeepPCF-MVS	3452.42 16	2670.71 24	3021.57 24	1618.12 25	2377.67 28	1751.39 26	8915.70 26	567.44 24	4569.74 24	1236.84 21	4615.63 25	513.23 24	2291.32 25	3515.48 23	882.39 23	811.66 18

LF4IMVS		2699.50 25	3049.67 25	1649.00 26	1755.00 23	1689.00 25	8330.00 23	464.00 20	4997.00 27	1528.00 26	4763.00 27	568.00 26	2296.00 26	4215.00 28	827.00 20	962.00 24

3Dnovator+	3414.41 15	3146.67 26	3537.00 26	1975.67 28	2310.00 27	2005.00 29	10100.00 27	626.00 26	5753.00 28	1803.00 29	5179.00 28	487.00 23	2622.00 28	4550.00 29	1025.00 28	1300.00 30

DeepC-MVS	3917.13 17	3155.06 27	3601.83 27	1814.72 27	2515.52 29	1941.53 27	10432.50 28	589.42 25	4789.45 25	1298.65 23	7677.31 33	578.35 27	2589.53 27	3611.78 25	923.52 24	913.10 23
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
TAPA-MVS	6309.84 19	3374.90 28	3806.99 28	2078.63 30	2535.47 30	2499.25 30	8816.16 25	456.05 19	6255.60 30	1664.52 28	6704.88 30	629.95 28	2768.38 29	6343.93 32	856.35 22	968.25 25
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
MAR-MVS		3973.16 29	4387.92 29	2728.86 31	3345.27 31	3341.39 34	11403.90 29	798.76 29	6555.62 33	2569.29 31	6514.86 29	834.97 34	3237.43 31	5891.08 31	1577.54 32	1607.76 31
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
PHI-MVS		4342.33 30	4863.11 30	2780.00 32	3743.00 32	3098.00 31	13388.00 30	897.00 32	6476.00 31	2782.00 32	7237.00 31	752.00 30	3591.00 32	6599.00 33	1894.00 34	1651.00 34

0108		4380.83 31	4906.56 31	2803.67 33	3844.00 33	3116.00 33	13454.00 31	901.00 33	6511.00 32	2794.00 33	7365.00 32	754.00 31	3634.00 33	6622.00 34	1914.00 35	1661.00 35

OMC-MVS		5022.26 32	5748.82 32	2842.58 34	4905.27 35	3101.27 32	14933.60 32	812.24 30	9278.40 34	2959.29 34	8800.17 34	1092.24 38	3806.30 34	7568.54 36	1389.60 31	1620.18 33

CLD-MVS		7338.25 33	8341.00 33	4330.00 35	4902.00 34	4991.00 35	23318.00 34	728.00 28	14437.00 37	4378.00 36	13379.00 35	824.00 33	5708.00 35	11429.00 38	1674.00 33	2291.00 36
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
LS3D		7684.71 34	8776.23 34	4410.16 36	5298.55 36	5631.53 36	34329.90 36	1504.57 36	11554.00 36	3896.41 35	15111.00 36	812.19 32	5985.63 36	3516.20 24	2963.23 36	1613.31 32

MVE	10021.00 20	10550.67 35	10002.44 35	12195.33 38	6242.00 37	8768.00 38	30896.00 35	4451.00 38	10331.00 35	6247.00 37	18296.00 37	283.00 15	21484.00 38	7364.00 35	5912.00 38	6334.00 38
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
CNLPA		10738.67 36	11772.44 36	7637.33 37	9120.00 38	8206.00 37	22252.00 33	2010.00 37	24984.00 38	7458.00 38	22596.00 38	2940.00 39	10260.00 37	11100.00 37	3492.00 37	4446.00 37

LTVRE_ROB	20565.47 21	23378.08 37	22805.00 37	25097.33 39	15500.00 39	19201.00 39	42712.00 37	6055.00 39	38388.00 39	16926.00 39	38534.00 39	7071.00 40	35653.00 39	29475.00 39	10584.00 39	20438.00 40
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
MSDG		32606.17 38	34443.44 38	27094.33 40	29249.00 40	25246.00 40	51144.00 38	6705.00 40	42886.00 40	22888.00 40	48534.00 40	13263.00 41	37507.00 40	79298.00 40	16024.00 40	18530.00 39

wuykxyi23d		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

PAPM		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

wuyk23d		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

PNet_i23d		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

TDRefinement		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

ANet_high		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

ACMMwithACMP		10000000.00 39	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 39	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 42	10000000.00 41	10000000.00 41	10000000.00 41	10000000.00 41

DeepMVS_CX				1.00 1	1.00 1	1.00 1		1.00 1	1.00 1	1.00 1	1.00 1	1.00 1	1.00 1	1.00 1	1.00 1	1.00 1

ITE_SJBPF				2014.00 29	1998.00 26	1992.00 28		701.00 27	5989.00 29	1812.00 30	3612.00 21	650.00 29	2852.00 30	4601.00 30	1195.00 30	1198.00 29

Test By Simon												990.00 35

High-resolution multi-view benchmark