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.
sort bysorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
LTVRE_ROB86.10 193.04 193.44 191.82 293.73 385.72 396.79 195.51 188.86 395.63 196.99 184.81 593.16 391.10 197.53 596.58 1
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_ROBcopyleft83.01 291.97 291.95 292.04 193.68 486.15 293.37 395.10 290.28 292.11 495.03 489.75 494.93 279.95 598.27 395.04 2
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
ACMM90.65 390.99 489.63 495.03 283.53 689.62 593.35 579.20 793.83 393.60 790.81 292.96 485.02 298.45 192.41 6
LS3D90.60 490.34 691.38 389.03 784.23 593.58 294.68 390.65 190.33 593.95 684.50 695.37 180.87 495.50 894.53 3
PMVScopyleft80.48 390.08 590.66 588.34 596.71 192.97 190.31 489.57 888.51 490.11 695.12 390.98 188.92 677.55 697.07 683.13 10
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
ACMH76.49 489.34 691.14 383.96 792.50 570.36 889.55 693.84 481.89 694.70 295.44 290.69 388.31 883.33 398.30 293.20 5
Gipumacopyleft84.44 786.33 778.78 1084.20 1073.57 789.55 690.44 784.24 584.38 894.89 576.35 980.40 1176.14 796.80 782.36 11
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
MVELAS83.18 882.64 884.79 689.05 667.82 1077.93 1092.52 668.33 1085.07 781.54 1282.06 792.96 469.35 897.91 493.57 4
MSDG80.06 979.99 980.25 883.91 1168.04 977.51 1189.19 977.65 881.94 1083.45 1176.37 886.31 963.31 1086.59 1086.41 8
OpenMVS_ROBcopyleft70.19 577.77 1077.46 1078.71 1184.39 961.15 1281.18 982.52 1062.45 1183.34 987.37 1066.20 1088.66 764.69 985.02 1186.32 9
CMPMVSbinary59.41 675.12 1173.57 1179.77 975.84 1267.22 1181.21 882.18 1150.78 1276.50 1287.66 855.20 1282.99 1062.17 1190.64 989.09 7
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
MVEpermissive40.22 751.82 1250.47 1255.87 1362.66 1351.91 1331.61 1239.28 1340.65 1350.76 1374.98 1356.24 1144.67 1333.94 1364.11 1371.04 12
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
FPMVS73.14 1288.63 885.00 467.39 1271.94 977.80 1187.66 850.48 1375.83 1249.95 1279.51 1258.58 13