This table lists the benchmark results for the high-res multi-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 Infoallhigh-res
multi-view
indooroutdoorbotani.boulde.bridgedoorexhibi.lectur.living.loungeobserv.old co.statueterrac.
sorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
LTVRE_ROB69.57 176.25 474.54 481.41 588.60 464.38 779.24 989.12 370.76 269.79 387.86 349.09 493.20 456.21 480.16 586.65 5
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
ACMH67.68 275.89 573.93 581.77 488.71 366.61 488.62 389.01 469.81 466.78 586.70 541.95 791.51 655.64 578.14 687.17 4
COLMAP_ROBcopyleft66.92 373.01 770.41 780.81 687.13 665.63 688.30 484.19 762.96 663.80 887.69 438.04 892.56 546.66 774.91 984.24 8
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
OpenMVS_ROBcopyleft64.09 470.56 868.19 877.65 880.26 859.41 885.01 682.96 958.76 865.43 782.33 937.63 991.23 745.34 976.03 882.32 9
CMPMVSbinary51.72 570.19 968.16 976.28 973.15 1057.55 979.47 883.92 848.02 956.48 984.81 743.13 686.42 962.67 381.81 484.89 6
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
PMVScopyleft37.38 644.16 1140.28 1155.82 1140.82 1342.54 1265.12 1063.99 1234.43 1124.48 1257.12 113.92 1276.17 1017.10 1255.52 1248.75 11
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
MVEpermissive26.22 730.37 1225.89 1243.81 1344.55 1235.46 1328.87 1239.07 1318.20 1318.58 1340.18 122.68 1347.37 1317.07 1323.78 1348.60 12
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
MSDG73.36 670.99 680.49 784.51 765.80 580.71 786.13 665.70 565.46 683.74 844.60 590.91 851.13 676.89 784.74 7
ACMM80.78 279.84 283.58 289.31 268.37 189.99 191.60 270.28 377.25 289.66 253.37 293.53 374.24 282.85 388.85 2
LS3D76.95 374.82 383.37 390.45 167.36 389.15 286.94 561.87 769.52 490.61 151.71 394.53 146.38 886.71 288.21 3
MVELAS82.31 181.65 184.29 188.47 567.73 285.81 592.35 175.78 178.33 186.58 664.01 194.35 276.05 187.48 190.79 1
FPMVS59.81 1065.08 1151.03 1169.58 1141.46 1040.67 1072.32 1016.46 1070.00 1224.24 1065.42 1058.40 10
Gipumacopyleft45.18 1041.86 1055.16 1277.03 951.52 1032.50 1180.52 1032.46 1227.12 1135.02 139.52 1175.50 1122.31 1160.21 1138.45 13
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015