This table lists the benchmark results for the low-res many-view scenario. The following metrics are evaluated:

(*) 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 Infoalllow-res
many-view
indooroutdoorlakesidesand boxstorage roomstorage room 2tunnel
sort bysorted bysort bysort bysort bysort bysort bysort bysort by
DeepPCF-MVS96.37 292.61 189.11 194.94 195.58 195.07 185.62 192.60 194.17 1
DeepC-MVS_fast96.70 192.27 288.75 294.63 295.15 394.80 285.14 392.35 293.92 2
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
DeepC-MVS95.98 391.43 387.48 494.06 494.52 594.15 583.69 691.26 393.50 3
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
TAPA-MVS93.98 790.68 488.05 392.44 792.42 1093.52 885.24 290.85 691.36 14
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
TAPA-MVS(SR)90.57 587.48 492.63 692.64 893.17 1183.91 591.04 492.10 12
PCF-MVS93.45 1090.50 685.08 1494.11 395.54 293.44 1082.27 1287.88 1693.34 4
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
COLMAP(SR)90.17 785.79 1293.08 592.63 993.87 682.51 1189.07 1492.74 6
PLCcopyleft95.07 489.96 887.00 691.94 1290.65 1992.68 1683.01 890.98 592.48 7
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
COLMAP(base)89.91 986.37 992.28 991.56 1592.95 1382.53 1090.21 892.33 8
ACMH+92.99 1389.84 1085.96 1192.43 892.28 1192.84 1581.89 1390.03 992.17 10
GSE89.80 1186.71 791.87 1390.87 1892.62 1783.49 789.92 1092.11 11
ACMM93.85 889.42 1285.42 1392.10 1191.09 1792.98 1281.16 1589.68 1192.23 9
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
ACMP93.49 989.04 1384.30 1792.21 1090.34 2093.47 979.54 1989.07 1492.81 5
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
LTVRE_ROB92.95 1489.00 1486.32 1090.78 1991.67 1489.54 2584.87 487.77 1791.13 16
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
ACMH92.88 1588.89 1584.55 1591.78 1491.37 1692.41 1879.88 1789.21 1291.56 13
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
IB-MVS91.98 1688.88 1686.61 890.39 2092.23 1292.94 1482.93 990.30 786.00 23
COLMAP_ROBcopyleft93.27 1188.32 1784.41 1690.93 1889.38 2292.11 1979.72 1889.10 1391.28 15
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
3Dnovator94.51 587.43 1880.90 2291.78 1494.76 494.22 378.96 2082.85 2286.35 20
3Dnovator+94.38 687.07 1980.26 2391.61 1694.34 694.17 478.01 2182.52 2386.34 21
A-TVSNet + Gipumacopyleft86.96 2082.91 1989.65 2288.10 2491.71 2080.23 1685.60 2189.15 18
BP-MVSNet86.71 2181.68 2090.06 2191.85 1387.55 3075.75 2487.61 1890.77 17
Christian Sormann, Patrick Knöbelreiter, Andreas Kuhn, Mattia Rossi, Thomas Pock, Friedrich Fraundorfer: BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo. 3DV 2020
OpenMVScopyleft93.04 1286.34 2279.08 2491.18 1793.50 793.77 776.56 2281.60 2686.26 22
LPCS85.52 2383.61 1886.79 2484.99 3088.49 2781.34 1485.87 2086.89 19
PVSNet_088.72 1882.59 2475.14 2887.55 2388.79 2388.38 2873.35 2776.93 3085.48 24
PVSNet91.96 1781.42 2574.82 2985.83 2787.90 2586.77 3174.82 2574.81 3482.83 25
R-MVSNet81.26 2677.44 2583.81 2982.82 3186.25 3274.52 2680.36 2882.37 26
CIDER80.48 2772.14 3186.05 2587.66 2689.77 2371.75 2872.53 3880.72 28
Qingshan Xu and Wenbing Tao: Learning Inverse Depth Regression for Multi-View Stereo with Correlation Cost Volume. AAAI 2020
test_120580.31 2875.46 2783.55 3086.48 2890.90 2170.38 3080.54 2773.26 39
test_112680.30 2971.93 3385.88 2689.75 2190.41 2270.18 3173.68 3677.50 32
OpenMVS_ROBcopyleft86.42 1980.25 3071.88 3485.82 2887.52 2789.56 2470.53 2973.24 3780.39 29
ANet-0.7577.73 3174.78 3079.70 3377.01 3784.71 3567.34 3582.21 2577.39 33
P-MVSNet77.69 3281.61 2175.07 3971.91 4377.33 4176.51 2386.71 1975.98 36
CPR_FA74.23 3372.06 3275.68 3774.30 4170.69 4768.60 3375.52 3282.05 27
ANet74.07 3468.26 3677.94 3677.01 3784.71 3564.75 3671.76 3972.09 40
AttMVS72.35 3568.13 3775.16 3870.98 4780.63 3768.09 3468.16 4073.87 38
Pnet-new-70.55 3653.66 4881.81 3281.27 3385.23 3451.15 4556.17 5478.92 31
CasMVSNet(SR_A)69.66 3755.07 4479.38 3480.08 3487.62 2946.56 5063.58 4770.44 41
Pnet_fast68.89 3847.01 5783.47 3181.84 3289.30 2638.61 5855.42 5579.28 30
CasMVSNet(base)68.63 3953.70 4778.58 3579.92 3585.96 3344.83 5262.57 4969.87 42
A1Net68.49 4075.64 2663.72 4963.53 5451.39 6268.88 3282.41 2476.23 35
MVSNet_plusplus64.09 4148.06 5574.77 4085.29 2962.00 5231.11 6265.01 4177.03 34
Pnet-blend++64.02 4252.13 5271.95 4175.81 3978.00 3940.66 5563.61 4562.05 47
Pnet-blend64.02 4252.13 5271.95 4175.81 3978.00 3940.66 5563.61 4562.05 47
unMVSv163.91 4462.59 4064.80 4866.39 5068.00 5061.62 3763.55 4860.01 50
MVSNet63.58 4556.25 4268.47 4766.00 5171.12 4448.36 4764.13 4268.29 44
MVSCRF63.13 4652.16 5170.45 4370.18 4875.83 4351.17 4453.16 5865.33 45
RMVSNet62.69 4769.01 3558.48 5769.15 4958.56 5659.72 3878.30 2947.75 60
metmvs_fine60.95 4866.50 3857.25 5964.00 5251.92 6059.04 3973.97 3555.84 55
Snet60.77 4947.18 5669.82 4477.90 3661.73 5339.61 5754.76 5669.82 43
Pnet-eth60.43 5065.87 3956.80 6063.59 5345.39 6555.58 4176.16 3161.43 49
MVSNet_++59.91 5145.04 5869.82 4474.28 4260.08 5414.97 6575.11 3375.10 37
hgnet59.89 5254.16 4563.70 5071.39 4470.87 4552.09 4256.24 5248.85 57
DPSNet59.89 5254.16 4563.70 5071.39 4470.87 4552.09 4256.24 5248.85 57
MVEpermissive62.14 2159.41 5460.38 4158.77 5650.39 6369.00 4956.98 4063.78 4456.91 54
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
test_112458.90 5544.17 6068.72 4662.60 5978.44 3838.12 5950.22 5965.10 46
MVSNet + Gipuma57.07 5653.08 4959.73 5562.91 5757.42 5848.41 4657.74 5058.88 51
CasMVSNet(SR_B)56.87 5755.95 4357.48 5856.06 6067.85 5147.86 4964.05 4348.53 59
F/T MVSNet+Gipuma55.78 5849.83 5459.76 5463.10 5557.51 5746.34 5153.31 5758.66 52
example55.06 5944.75 5961.93 5271.21 4669.85 4842.82 5446.68 6044.73 62
firsttry54.62 6052.74 5055.88 6154.90 6155.69 5948.02 4857.46 5157.06 53
PMVScopyleft61.03 2249.24 6132.69 6560.27 5362.70 5876.26 4242.99 5322.38 6541.86 63
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
unMVSmet48.01 6240.73 6252.86 6253.90 6258.72 5536.72 6144.74 6245.96 61
confMetMVS44.56 6341.82 6146.39 6346.63 6451.78 6137.38 6046.25 6140.77 64
test_1120copyleft40.25 6435.81 6443.21 6430.29 6650.38 6329.89 6441.72 6448.95 56
Cas-MVS_preliminary38.91 6536.73 6340.37 6537.57 6548.76 6430.59 6342.87 6334.76 65
CMPMVSbinary66.06 209.32 660.44 6715.23 666.17 6739.51 660.89 670.00 670.00 67
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
FADENet4.02 674.30 663.83 676.03 683.71 675.02 663.58 661.74 66
dnet0.00 680.00 680.00 680.00 690.00 680.00 680.00 670.00 67
UnsupFinetunedMVSNet63.10 55