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 bysort bysort bysort bysort bysort bysort bysort bysort bysort bysort bysorted bysort bysort by
DeepPCF-MVS80.84 188.10 188.56 186.73 592.24 669.03 1089.57 1193.39 377.53 289.79 194.12 278.98 196.58 485.66 195.72 194.58 2
DeepC-MVS79.81 287.08 286.88 287.69 391.16 972.32 690.31 793.94 177.12 482.82 394.23 172.13 297.09 184.83 295.37 293.65 4
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
ACMMwithACMP85.89 485.39 487.38 493.59 172.63 392.74 293.18 476.78 580.73 593.82 364.33 796.29 582.67 390.69 993.23 5
DeepC-MVS_fast79.65 386.91 386.62 387.76 293.52 272.37 491.26 493.04 576.62 684.22 293.36 571.44 396.76 280.82 495.33 394.16 3
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020
ACMP74.13 681.51 1280.57 1284.36 1189.42 1468.69 1389.97 991.50 1074.46 1075.04 1690.41 1253.82 1894.54 1177.56 582.91 2089.86 12
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020
3Dnovator+77.84 485.48 584.47 588.51 191.08 1073.49 193.18 193.78 280.79 176.66 1393.37 460.40 1396.75 377.20 693.73 495.29 1
CLD-MVS82.31 1081.65 1084.29 1288.47 2167.73 1685.81 2392.35 675.78 778.33 886.58 2664.01 894.35 1476.05 787.48 1390.79 11
Zhaoxin Li, Wangmeng Zuo, Zhaoqi Wang, Lei Zhang: Confidence-based Large-scale Dense Multi-view Stereo. IEEE Transaction on Image Processing, 2020
OMC-MVS82.69 981.97 984.85 1088.75 1767.42 1787.98 2190.87 1474.92 979.72 691.65 962.19 1193.96 1575.26 886.42 1793.16 6
ACMM73.20 880.78 1379.84 1383.58 1589.31 1568.37 1489.99 891.60 870.28 1377.25 1189.66 1553.37 2093.53 1974.24 982.85 2188.85 16
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
3Dnovator76.31 583.38 782.31 786.59 687.94 2272.94 290.64 592.14 777.21 375.47 1592.83 658.56 1494.72 1073.24 1092.71 592.13 9
MG-MVS83.41 683.45 683.28 1792.74 462.28 2688.17 1989.50 1775.22 881.49 492.74 766.75 495.11 772.85 1191.58 792.45 7
PHI-MVScopyleft82.98 882.13 885.52 792.26 572.35 591.75 390.46 1574.46 1078.30 989.65 1664.71 694.98 967.62 1289.95 1089.22 14
MAR-MVS81.84 1180.70 1185.27 891.32 871.53 789.82 1090.92 1369.77 1678.50 786.21 2762.36 1094.52 1365.36 1392.05 689.77 13
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
AdaColmapbinary80.58 1479.42 1484.06 1493.09 368.91 1289.36 1288.97 2269.27 1775.70 1489.69 1457.20 1695.77 663.06 1488.41 1187.50 24
CMPMVSbinary51.72 1970.19 3268.16 3276.28 3173.15 3657.55 3179.47 3083.92 2948.02 3756.48 3584.81 2943.13 2786.42 3162.67 1581.81 2384.89 29
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011
OpenMVScopyleft72.83 1079.77 1678.33 1684.09 1385.17 2769.91 990.57 690.97 1266.70 1972.17 1891.91 854.70 1793.96 1561.81 1690.95 888.41 19
PCF-MVS73.52 780.38 1578.84 1585.01 987.71 2368.99 1183.65 2691.46 1163.00 2777.77 1090.28 1366.10 595.09 861.40 1788.22 1290.94 10
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019
CNLPA78.08 1876.79 1881.97 2290.40 1271.07 887.59 2284.55 2766.03 2072.38 1789.64 1757.56 1586.04 3259.61 1883.35 1988.79 17
PLCcopyleft70.83 1178.05 1976.37 2083.08 1891.88 767.80 1588.19 1889.46 1864.33 2469.87 2288.38 2053.66 1993.58 1858.86 1982.73 2287.86 21
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019
ACMH+68.96 1376.01 2474.01 2482.03 2188.60 1965.31 2388.86 1487.55 2370.25 1467.75 2587.47 2341.27 3093.19 2258.37 2075.94 3187.60 23
F-COLMAP76.38 2274.33 2382.50 2089.28 1666.95 1988.41 1689.03 2064.05 2666.83 2788.61 1946.78 2492.89 2357.48 2178.55 2587.67 22
PatchmatchNet73.12 2771.33 2678.49 2883.18 2960.85 2879.63 2978.57 3664.13 2571.73 1979.81 3551.20 2285.97 3357.40 2276.36 2888.66 18
LTVRE_ROB69.57 1276.25 2374.54 2281.41 2588.60 1964.38 2479.24 3189.12 1970.76 1269.79 2387.86 2149.09 2393.20 2156.21 2380.16 2486.65 28
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 1475.89 2573.93 2581.77 2388.71 1866.61 2088.62 1589.01 2169.81 1566.78 2886.70 2541.95 2991.51 2555.64 2478.14 2687.17 26
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019
PVSNet64.34 1672.08 2970.87 2875.69 3386.21 2656.44 3474.37 3480.73 3462.06 2970.17 2182.23 3342.86 2883.31 3454.77 2584.45 1887.32 25
ITE_SJBPF78.22 2981.77 3160.57 2983.30 3069.25 1867.54 2687.20 2436.33 3387.28 3054.34 2674.62 3386.80 27
USDC70.33 3168.37 3076.21 3280.60 3256.23 3579.19 3286.49 2560.89 3261.29 3285.47 2831.78 3489.47 2853.37 2776.21 2982.94 32
LF4IMVS64.02 3462.19 3469.50 3570.90 3753.29 3676.13 3377.18 3752.65 3658.59 3380.98 3423.55 3676.52 3653.06 2866.66 3578.68 36
PAPM77.68 2076.40 1981.51 2487.29 2461.85 2783.78 2589.59 1664.74 2371.23 2088.70 1862.59 993.66 1752.66 2987.03 1589.01 15
MSDG73.36 2670.99 2780.49 2784.51 2865.80 2180.71 2786.13 2665.70 2265.46 2983.74 3044.60 2590.91 2751.13 3076.89 2784.74 30
TAPA-MVS73.13 979.15 1777.94 1782.79 1989.59 1362.99 2588.16 2091.51 965.77 2177.14 1291.09 1060.91 1293.21 2050.26 3187.05 1492.17 8
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019
PVSNet_057.27 1861.67 3559.27 3568.85 3679.61 3457.44 3268.01 3773.44 3855.93 3558.54 3470.41 3744.58 2677.55 3547.01 3235.91 3971.55 37
COLMAP_ROBcopyleft66.92 1573.01 2870.41 2980.81 2687.13 2565.63 2288.30 1784.19 2862.96 2863.80 3187.69 2238.04 3192.56 2446.66 3374.91 3284.24 31
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016
LS3D76.95 2174.82 2183.37 1690.45 1167.36 1889.15 1386.94 2461.87 3069.52 2490.61 1151.71 2194.53 1246.38 3486.71 1688.21 20
TinyColmap67.30 3364.81 3374.76 3481.92 3056.68 3380.29 2881.49 3260.33 3356.27 3683.22 3124.77 3587.66 2945.52 3569.47 3479.95 34
OpenMVS_ROBcopyleft64.09 1770.56 3068.19 3177.65 3080.26 3359.41 3085.01 2482.96 3158.76 3465.43 3082.33 3237.63 3291.23 2645.34 3676.03 3082.32 33
wuykxyi23d39.76 4033.18 4159.51 3946.98 4244.01 3957.70 4067.74 4024.13 4213.98 4434.33 441.27 4471.33 4034.23 3718.23 4263.18 38
ANet_high50.57 3746.10 3763.99 3748.67 4139.13 4170.99 3580.85 3361.39 3131.18 3857.70 3817.02 3773.65 3931.22 3815.89 4479.18 35
PNet_i23d38.26 4135.42 4046.79 4258.74 3935.48 4259.65 3951.25 4232.45 4123.44 4247.53 412.04 4358.96 4225.60 3918.09 4345.92 42
FPMVS53.68 3651.64 3659.81 3865.08 3851.03 3869.48 3669.58 3941.46 3840.67 3772.32 3616.46 3870.00 4124.24 4065.42 3658.40 39
Gipumacopyleft45.18 3841.86 3855.16 4177.03 3551.52 3732.50 4280.52 3532.46 4027.12 3935.02 439.52 4075.50 3822.31 4160.21 3738.45 43
S. Galliani, K. Lasinger, K. Schindler: Massively Parallel Multiview Stereopsis by Surface Normal Diffusion. ICCV 2015
DeepMVS_CXcopyleft27.40 4440.17 4526.90 4524.59 4417.44 4423.95 4148.61 409.77 3926.48 4418.06 4224.47 4028.83 44
PMVScopyleft37.38 2044.16 3940.28 3955.82 4040.82 4442.54 4065.12 3863.99 4134.43 3924.48 4057.12 393.92 4176.17 3717.10 4355.52 3848.75 40
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010)
MVEpermissive26.22 2130.37 4225.89 4243.81 4344.55 4335.46 4328.87 4339.07 4318.20 4318.58 4340.18 422.68 4247.37 4317.07 4423.78 4148.60 41
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014)
wuyk23d16.82 4315.94 4319.46 4558.74 3931.45 4439.22 413.74 456.84 456.04 452.70 451.27 4424.29 4510.54 4514.40 452.63 45