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 bysort bysort bysorted bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort by
3Dnovator+77.84 485.48 484.47 488.51 191.08 873.49 193.18 193.78 180.79 176.66 1093.37 360.40 996.75 377.20 593.73 495.29 1
DeepC-MVS_fast79.62 286.82 286.47 387.85 293.71 172.56 391.82 293.01 477.89 283.06 393.31 470.46 396.87 279.94 395.02 294.14 3
DeepC-MVS79.50 386.80 386.53 287.61 391.96 572.46 490.85 393.50 276.69 583.93 293.40 272.49 296.91 181.69 294.27 393.44 4
DeepPCF-MVS80.84 188.10 188.56 186.73 492.24 469.03 789.57 993.39 377.53 389.79 194.12 178.98 196.58 485.66 195.72 194.58 2
3Dnovator76.31 583.38 682.31 686.59 587.94 1772.94 290.64 492.14 677.21 475.47 1292.83 558.56 1094.72 873.24 892.71 592.13 7
MAR-MVS81.84 880.70 885.27 691.32 771.53 589.82 890.92 1269.77 1378.50 586.21 2162.36 794.52 1165.36 1092.05 689.77 11
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
PCF-MVS73.52 780.38 1278.84 1285.01 787.71 1868.99 883.65 2091.46 1063.00 2077.77 790.28 1166.10 595.09 761.40 1488.22 1090.94 8
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
ACMP74.13 681.51 980.57 984.36 889.42 1168.69 1089.97 791.50 974.46 875.04 1390.41 1053.82 1394.54 977.56 482.91 1489.86 10
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
CLD-MVS82.31 781.65 784.29 988.47 1667.73 1385.81 1892.35 575.78 678.33 686.58 2064.01 694.35 1276.05 687.48 1190.79 9
OpenMVScopyleft72.83 1079.77 1378.33 1384.09 1085.17 2069.91 690.57 590.97 1166.70 1672.17 1491.91 754.70 1293.96 1361.81 1390.95 888.41 13
AdaColmapbinary80.58 1179.42 1184.06 1193.09 268.91 989.36 1088.97 1769.27 1475.70 1189.69 1257.20 1195.77 563.06 1188.41 987.50 17
ACMM73.20 880.78 1079.84 1083.58 1289.31 1268.37 1189.99 691.60 770.28 1077.25 889.66 1353.37 1593.53 1574.24 782.85 1588.85 12
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
LS3D76.95 1674.82 1683.37 1390.45 967.36 1489.15 1186.94 1961.87 2269.52 1790.61 951.71 1694.53 1046.38 2486.71 1388.21 14
MG-MVS83.41 583.45 583.28 1492.74 362.28 2188.17 1689.50 1375.22 781.49 492.74 666.75 495.11 672.85 991.58 792.45 5
PLCcopyleft70.83 1178.05 1576.37 1583.08 1591.88 667.80 1288.19 1589.46 1464.33 1969.87 1588.38 1453.66 1493.58 1458.86 1582.73 1687.86 15
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
TAPA-MVS73.13 979.15 1477.94 1482.79 1689.59 1062.99 2088.16 1791.51 865.77 1777.14 991.09 860.91 893.21 1650.26 2287.05 1292.17 6
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
ACMH+68.96 1376.01 1874.01 1882.03 1788.60 1465.31 1888.86 1287.55 1870.25 1167.75 1887.47 1741.27 2193.19 1858.37 1675.94 2287.60 16
ACMH67.68 1475.89 1973.93 1981.77 1888.71 1366.61 1588.62 1389.01 1669.81 1266.78 2086.70 1941.95 2091.51 2055.64 1878.14 1987.17 18
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
LTVRE_ROB69.57 1276.25 1774.54 1781.41 1988.60 1464.38 1979.24 2389.12 1570.76 969.79 1687.86 1549.09 1793.20 1756.21 1780.16 1886.65 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_ROBcopyleft66.92 1573.01 2170.41 2180.81 2087.13 1965.63 1788.30 1484.19 2162.96 2163.80 2387.69 1638.04 2292.56 1946.66 2374.91 2384.24 23
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
MSDG73.36 2070.99 2080.49 2184.51 2165.80 1680.71 2186.13 2065.70 1865.46 2183.74 2344.60 1890.91 2251.13 2176.89 2084.74 22
ITE_SJBPF78.22 2281.77 2260.57 2283.30 2369.25 1567.54 1987.20 1836.33 2487.28 2354.34 1974.62 2486.80 19
OpenMVS_ROBcopyleft64.09 1670.56 2268.19 2277.65 2380.26 2359.41 2385.01 1982.96 2458.76 2465.43 2282.33 2437.63 2391.23 2145.34 2576.03 2182.32 24
CMPMVSbinary51.72 1770.19 2368.16 2376.28 2473.15 2557.55 2479.47 2283.92 2248.02 2656.48 2584.81 2243.13 1986.42 2462.67 1281.81 1784.89 21
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
LF4IMVS64.02 2462.19 2469.50 2570.90 2653.29 2576.13 2477.18 2752.65 2558.59 2480.98 2523.55 2576.52 2553.06 2066.66 2578.68 26
ANet_high50.57 2546.10 2563.99 2648.67 3039.13 3070.99 2580.85 2561.39 2331.18 2757.70 2717.02 2673.65 2831.22 2715.89 3379.18 25
FPMVS59.81 2765.08 2751.03 2769.58 2841.46 2740.67 2672.32 2616.46 2770.00 3024.24 2965.42 2658.40 28
wuykxyi23d39.76 2833.18 2959.51 2846.98 3144.01 2857.70 2867.74 2924.13 3113.98 3334.33 331.27 3371.33 2934.23 2618.23 3163.18 27
PMVScopyleft37.38 1844.16 2740.28 2755.82 2940.82 3342.54 2965.12 2663.99 3034.43 2824.48 2957.12 283.92 3076.17 2617.10 3255.52 2848.75 29
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
Gipumacopyleft45.18 2641.86 2655.16 3077.03 2451.52 2632.50 3080.52 2632.46 2927.12 2835.02 329.52 2975.50 2722.31 3060.21 2738.45 32
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
PNet_i23d38.26 2935.42 2846.79 3158.74 2835.48 3159.65 2751.25 3132.45 3023.44 3147.53 302.04 3258.96 3125.60 2818.09 3245.92 31
MVEpermissive26.22 1930.37 3025.89 3043.81 3244.55 3235.46 3228.87 3139.07 3218.20 3218.58 3240.18 312.68 3147.37 3217.07 3323.78 3048.60 30
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
DeepMVS_CXcopyleft27.40 3340.17 3426.90 3424.59 3317.44 3323.95 3048.61 299.77 2826.48 3318.06 3124.47 2928.83 33
wuyk23d16.82 3115.94 3119.46 3458.74 2831.45 3339.22 293.74 346.84 346.04 342.70 341.27 3324.29 3410.54 3414.40 342.63 34