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 bysort bysorted bysort bysort bysort bysort bysort bysort by
DeepPCF-MVS80.84 163.41 153.95 169.71 269.72 471.26 148.73 159.18 168.16 2
DeepC-MVS79.81 262.37 252.77 268.77 368.65 570.30 346.95 258.59 367.37 3
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
TAPA-MVS73.13 958.67 652.34 362.89 1163.68 1063.46 1145.85 458.83 261.53 11
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
DeepC-MVS_fast79.65 362.24 350.96 469.75 169.91 371.08 246.04 355.88 468.27 1
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
3Dnovator76.31 559.36 548.19 566.81 570.04 267.59 544.27 552.10 762.80 10
PCF-MVS73.52 757.06 748.10 663.03 966.85 762.36 1343.32 852.89 559.86 14
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
3Dnovator+77.84 459.53 447.18 767.76 470.63 167.68 443.34 751.01 964.98 6
PLCcopyleft70.83 1153.50 1545.79 858.64 1756.31 2660.76 1641.70 1149.88 1158.86 17
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
OpenMVScopyleft72.83 1056.18 1045.66 963.19 864.67 865.13 839.60 1451.71 859.76 15
LTVRE_ROB69.57 1353.52 1445.46 1058.89 1558.76 1960.60 1843.91 647.01 1557.32 20
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
COLMAP(base)55.51 1144.97 1162.54 1360.85 1562.46 1242.24 947.71 1364.29 7
ACMP74.13 656.41 844.60 1264.28 661.46 1465.81 737.07 1852.13 665.56 5
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
COLMAP(SR)56.21 944.16 1364.25 763.87 962.07 1441.86 1046.45 1766.80 4
ACMH+68.96 1452.02 1843.83 1457.49 2059.85 1651.27 3540.04 1247.62 1461.35 12
TAPA-MVS(SR)55.07 1243.33 1562.90 1067.41 657.81 2239.94 1346.71 1663.49 8
ACMM73.20 855.01 1343.19 1662.89 1159.60 1766.07 635.89 2150.48 1063.01 9
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
COLMAP_ROBcopyleft66.92 1752.32 1742.45 1758.89 1556.18 2761.09 1538.61 1546.28 1859.41 16
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
IB-MVS68.01 1549.19 2142.00 1853.98 2858.50 2051.98 3235.85 2248.16 1251.46 28
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
GSE52.52 1640.81 1960.34 1461.71 1358.23 2138.13 1743.48 2161.07 13
LPCS50.60 1940.43 2057.38 2158.37 2156.00 2638.14 1642.71 2357.77 18
HY-MVS69.67 1249.80 2040.39 2156.07 2459.17 1853.68 3036.13 1944.66 1955.36 23
AttMVS45.85 2639.26 2250.25 3649.36 3751.75 3434.83 2343.70 2049.63 32
OpenMVS_ROBcopyleft64.09 1948.56 2238.68 2355.15 2561.90 1249.65 3736.02 2041.33 2553.89 25
CIDER47.21 2438.36 2453.11 3055.37 2955.10 2933.25 2743.46 2248.87 34
Qingshan Xu and Wenbing Tao: Learning Inverse Depth Regression for Multi-View Stereo with Correlation Cost Volume. AAAI 2020
ACMH67.68 1647.97 2338.24 2554.45 2756.88 2350.08 3634.49 2441.99 2456.40 21
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
P-MVSNet44.46 3137.09 2649.37 3849.27 3849.30 3834.35 2539.83 2649.54 33
PVSNet64.34 1845.78 2736.40 2752.03 3152.49 3149.20 3934.18 2638.61 2854.41 24
PVSNet_057.27 2046.73 2534.96 2854.57 2655.50 2851.93 3331.22 2838.71 2756.27 22
BP-MVSNet43.22 3232.08 2950.65 3552.86 3046.16 4527.25 3036.92 2952.94 26
Christian Sormann, Patrick Knöbelreiter, Andreas Kuhn, Mattia Rossi, Thomas Pock, Friedrich Fraundorfer: BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo. 3DV 2020
R-MVSNet36.87 4129.78 3041.61 4442.00 4446.99 4124.73 3134.83 3035.83 44
test_120540.46 3629.35 3147.87 3947.88 4056.48 2524.37 3234.32 3139.23 43
A-TVSNet + Gipumacopyleft42.12 3529.09 3250.80 3348.28 3956.53 2429.14 2929.04 3947.59 35
test_112645.72 2828.57 3357.15 2262.33 1159.06 1924.22 3332.91 3550.08 30
mvs_zhu_103042.98 3327.14 3453.53 2952.24 3258.28 2020.89 3633.39 3350.07 31
MVSNet38.33 3926.99 3545.89 4040.49 4557.57 2320.52 3833.47 3239.63 42
CasMVSNet(SR_B)32.16 4226.80 3635.73 4734.57 4746.99 4120.55 3733.05 3425.63 52
CasMVSNet(SR_A)45.13 2925.99 3757.89 1856.32 2564.81 919.76 4032.23 3652.54 27
CasMVSNet(base)44.49 3024.93 3857.53 1956.38 2464.76 1018.64 4231.23 3851.43 29
ANet39.53 3724.60 3949.48 3749.42 3655.75 2722.08 3527.13 4243.25 38
A1Net23.74 5123.95 4023.61 5321.43 5619.62 6519.93 3927.96 4029.78 48
CPR_FA23.28 5223.62 4123.05 5419.33 5823.00 5819.34 4127.89 4126.81 51
ANet-0.7539.47 3822.61 4250.71 3449.43 3555.75 2718.21 4327.01 4346.96 36
Pnet-new-42.61 3422.17 4356.24 2358.21 2253.03 3124.08 3420.25 5357.48 19
unsupervisedMVS_cas30.64 4620.24 4437.57 4635.93 4642.74 4815.36 4525.12 4534.04 45
Pnet-eth16.08 6718.30 4514.59 6921.78 537.07 759.66 6726.95 4414.91 66
MVSCRF28.32 4818.18 4635.09 4832.16 5044.37 4714.66 4821.69 5028.73 49
test320.33 5417.79 4722.02 5517.88 6024.31 5514.91 4620.68 5223.87 53
Pnet_fast37.64 4017.70 4850.93 3249.72 3460.73 1710.59 6624.81 4642.34 39
Snet23.85 5017.41 4928.15 5032.72 4924.90 5412.22 5922.60 4926.83 50
MVSNet_++25.72 4917.38 5031.27 4933.47 4818.28 682.87 7631.90 3742.06 40
TVSNet20.06 5517.25 5121.93 5618.25 5923.77 5613.74 5320.77 5123.78 54
MVEpermissive26.22 2316.26 6616.97 5215.79 6811.75 7119.40 6614.45 4919.48 5516.21 64
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
SGNet18.50 5816.17 5320.06 6016.84 6322.87 5913.42 5418.92 5720.47 61
PSD-MVSNet17.93 6115.59 5419.48 6116.64 6622.17 6213.02 5518.15 5919.64 62
RMVSNet11.09 7115.49 558.17 759.16 7210.05 7212.57 5818.40 585.29 75
QQQNet18.88 5715.19 5621.34 5717.26 6223.30 5713.79 5216.60 6023.48 55
Pnet-blend31.51 4415.02 5742.51 4247.41 4146.50 436.99 7023.05 4733.62 46
Pnet-blend++31.51 4415.02 5742.51 4247.41 4146.50 436.99 7023.05 4733.62 46
F/T MVSNet+Gipuma17.31 6214.39 5919.26 6316.83 6419.67 6414.03 5014.75 6321.27 59
metmvs_fine10.03 7214.31 607.18 766.75 766.98 769.52 6819.10 567.81 70
ternet18.01 5914.28 6120.49 5815.75 6822.30 6011.95 6116.60 6023.43 57
SVVNet18.01 5914.28 6120.49 5815.75 6822.30 6011.95 6116.60 6023.43 57
MVSNet + Gipuma16.91 6314.12 6318.77 6616.39 6719.13 6713.98 5114.27 6420.78 60
test_112431.65 4313.28 6443.90 4144.21 4347.60 4014.80 4711.75 7039.88 41
MVSNet_plusplus28.64 4712.93 6539.11 4549.80 3321.55 636.17 7219.69 5445.98 37
firsttry15.88 6812.57 6618.08 6717.80 6117.14 6911.54 6313.61 6519.31 63
DPSNet16.45 6412.31 6719.21 6421.51 5430.31 5112.80 5611.82 685.81 72
hgnet16.45 6412.31 6719.21 6421.51 5430.31 5112.80 5611.82 685.81 72
CCVNet19.37 5612.14 6924.19 5220.50 5728.61 5311.03 6513.25 6623.48 55
unMVSv113.05 7012.03 7013.73 7014.18 7016.18 7012.13 6011.93 6710.84 69
PMVScopyleft37.38 2221.09 5311.49 7127.48 5124.09 5144.44 4615.98 447.01 7213.92 68
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
example15.69 6910.14 7219.39 6223.73 5230.32 5011.10 649.17 714.13 76
test_1120copyleft9.53 736.82 7311.33 728.74 739.06 737.57 696.06 7316.20 65
unMVSmet7.42 745.82 748.48 747.60 7411.29 715.84 735.81 746.56 71
Cas-MVS_preliminary7.24 763.51 759.72 737.03 757.46 744.87 742.16 7614.65 67
confMetMVS4.86 773.39 765.85 775.59 776.61 772.94 753.83 755.35 74
FADENet0.17 780.15 770.18 780.31 790.16 780.19 770.10 770.08 77
CMPMVSbinary51.72 217.38 750.03 7812.27 712.37 7834.46 490.06 780.00 780.00 78
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
dnet0.00 790.00 790.00 790.00 800.00 790.00 790.00 780.00 78
UnsupFinetunedMVSNet16.83 64