BP-MVSNet | | | 64.34 1 | 50.31 8 | 73.69 1 | 74.99 1 | 71.90 9 | 46.42 11 | 54.19 8 | 74.18 1 |
Christian Sormann, Patrick Knöbelreiter, Andreas Kuhn, Mattia Rossi, Thomas Pock, Friedrich Fraundorfer: BP-MVSNet: Belief-Propagation-Layers for Multi-View-Stereo. 3DV 2020 |
A1Net | | | 62.59 2 | 54.91 3 | 67.70 7 | 68.95 5 | 72.10 7 | 52.15 2 | 57.67 3 | 62.06 11 |
|
ACMP | | 71.68 8 | 62.33 3 | 51.03 7 | 69.86 4 | 66.41 7 | 74.93 1 | 47.36 7 | 54.70 7 | 68.24 2 |
Qingshan Xu and Wenbing Tao: Planar Prior Assisted PatchMatch Multi-View Stereo. AAAI 2020 |
A-TVSNet + Gipuma |  | | 61.02 4 | 56.34 1 | 64.14 13 | 64.32 10 | 66.91 21 | 52.09 3 | 60.60 1 | 61.19 14 |
|
COLMAP(SR) | | | 60.84 5 | 47.20 11 | 69.93 3 | 69.97 4 | 72.99 4 | 47.93 6 | 46.47 14 | 66.85 3 |
|
DeepPCF-MVS | | 81.17 1 | 60.64 6 | 53.51 5 | 65.40 12 | 64.27 11 | 69.40 16 | 51.28 5 | 55.74 6 | 62.53 10 |
|
ACMH+ | | 65.35 14 | 59.83 7 | 46.18 14 | 68.93 5 | 65.70 8 | 74.24 3 | 46.34 12 | 46.01 16 | 66.83 4 |
|
DeepC-MVS | | 77.85 3 | 59.42 8 | 49.21 9 | 66.23 11 | 64.01 12 | 70.42 13 | 46.20 14 | 52.23 9 | 64.26 7 |
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020 |
DeepC-MVS_fast | | 79.48 2 | 59.14 9 | 48.44 10 | 66.27 9 | 64.81 9 | 70.57 11 | 46.73 10 | 50.15 10 | 63.44 8 |
Andreas Kuhn, Christian Sormann, Mattia Rossi, Oliver Erdler, Friedrich Fraundorfer: DeepC-MVS: Deep Confidence Prediction for Multi-View Stereo Reconstruction. 3DV 2020 |
PVSNet_0 | | 68.08 13 | 59.14 9 | 42.70 18 | 70.10 2 | 74.33 2 | 74.52 2 | 38.12 20 | 47.28 13 | 61.46 13 |
|
TAPA-MVS | | 70.22 10 | 58.89 11 | 51.21 6 | 64.01 14 | 62.59 14 | 66.45 23 | 46.31 13 | 56.11 5 | 62.99 9 |
Andrea Romanoni, Matteo Matteucci: TAPA-MVS: Textureless-Aware PAtchMatch Multi-View Stereo. ICCV 2019 |
PCF-MVS | | 73.15 7 | 58.42 12 | 54.11 4 | 61.29 18 | 61.25 17 | 64.94 25 | 51.80 4 | 56.43 4 | 57.66 18 |
Andreas Kuhn, Shan Lin, Oliver Erdler: Plane Completion and Filtering for Multi-View Stereo Reconstruction. GCPR 2019 |
ACMM | | 69.62 11 | 58.27 13 | 46.31 13 | 66.24 10 | 60.67 19 | 72.46 5 | 42.63 16 | 50.00 11 | 65.60 6 |
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019 |
PVSNet | | 73.49 6 | 57.76 14 | 43.28 17 | 67.41 8 | 70.29 3 | 72.10 7 | 40.29 18 | 46.26 15 | 59.84 15 |
|
TAPA-MVS(SR) | | | 55.57 15 | 35.84 27 | 68.73 6 | 67.52 6 | 72.24 6 | 33.19 28 | 38.50 24 | 66.42 5 |
|
COLMAP(base) | | | 54.01 16 | 44.78 15 | 60.17 19 | 58.77 23 | 63.60 27 | 45.57 15 | 43.99 18 | 58.14 17 |
|
PLC |  | 68.80 12 | 53.38 17 | 47.19 12 | 57.51 25 | 57.17 24 | 61.22 31 | 46.82 9 | 47.56 12 | 54.14 22 |
Jie Liao, Yanping Fu, Qingan Yan, Chunxia xiao: Pyramid Multi-View Stereo with Local Consistency. Pacific Graphics 2019 |
3Dnovator+ | | 73.60 5 | 53.20 18 | 38.24 22 | 63.18 16 | 62.18 15 | 70.72 10 | 36.81 21 | 39.68 23 | 56.64 20 |
|
3Dnovator | | 73.91 4 | 52.81 19 | 38.17 23 | 62.58 17 | 60.95 18 | 69.63 15 | 35.98 25 | 40.36 22 | 57.16 19 |
|
ANet-0.75 | | | 52.71 20 | 55.63 2 | 50.76 36 | 44.38 38 | 59.75 32 | 53.05 1 | 58.21 2 | 48.17 29 |
|
ACMH | | 63.93 15 | 52.68 21 | 35.79 28 | 63.93 15 | 60.12 21 | 70.07 14 | 34.99 26 | 36.59 29 | 61.61 12 |
Qingshan Xu and Wenbing Tao: Multi-Scale Geometric Consistency Guided Multi-View Stereo. CVPR 2019 |
GSE | | | 52.65 22 | 41.64 19 | 59.98 21 | 55.71 27 | 65.76 24 | 41.60 17 | 41.69 20 | 58.48 16 |
|
OpenMVS |  | 70.45 9 | 51.01 23 | 37.34 24 | 60.13 20 | 56.52 26 | 67.91 18 | 33.72 27 | 40.96 21 | 55.96 21 |
|
CIDER | | | 49.33 24 | 33.81 29 | 59.69 23 | 61.47 16 | 68.26 17 | 31.50 29 | 36.11 30 | 49.34 28 |
Qingshan Xu and Wenbing Tao: Learning Inverse Depth Regression for Multi-View Stereo with Correlation Cost Volume. AAAI 2020 |
OpenMVS_ROB |  | 61.12 16 | 48.48 25 | 31.63 30 | 59.71 22 | 59.03 22 | 67.15 19 | 31.08 30 | 32.18 33 | 52.95 25 |
|
LPCS | | | 48.13 26 | 38.62 21 | 54.48 30 | 51.49 31 | 57.98 34 | 38.93 19 | 38.30 27 | 53.97 23 |
|
CPR_FA | | | 47.15 27 | 40.41 20 | 51.65 35 | 47.19 35 | 53.80 40 | 36.70 23 | 44.12 17 | 53.96 24 |
|
COLMAP_ROB |  | 57.96 18 | 45.89 28 | 37.03 26 | 51.79 34 | 49.43 34 | 56.42 39 | 36.78 22 | 37.28 28 | 49.53 27 |
Johannes L. Schönberger, Enliang Zheng, Marc Pollefeys, Jan-Michael Frahm: Pixelwise View Selection for Unstructured Multi-View Stereo. ECCV 2016 |
test_1205 | | | 45.54 29 | 30.23 31 | 55.74 27 | 56.56 25 | 66.59 22 | 21.99 34 | 38.47 25 | 44.08 36 |
|
LTVRE_ROB | | 59.60 17 | 43.60 30 | 37.31 25 | 47.80 37 | 45.94 37 | 49.52 45 | 36.19 24 | 38.43 26 | 47.93 30 |
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 |
unMVSv1 | | | 43.21 31 | 44.62 16 | 42.27 42 | 40.66 41 | 51.71 43 | 47.08 8 | 42.16 19 | 34.44 43 |
|
Snet | | | 41.99 32 | 16.05 47 | 59.29 24 | 60.27 20 | 70.55 12 | 9.62 53 | 22.48 45 | 47.04 32 |
|
CasMVSNet(SR_A) | | | 41.31 33 | 18.44 43 | 56.56 26 | 55.29 29 | 67.08 20 | 13.89 44 | 22.98 43 | 47.30 31 |
|
MVSNet_++ | | | 41.03 34 | 19.38 40 | 55.47 28 | 63.04 13 | 57.24 35 | 6.23 58 | 32.52 32 | 46.14 33 |
|
test_1126 | | | 40.23 35 | 19.47 39 | 54.08 31 | 55.58 28 | 62.13 30 | 15.66 40 | 23.27 42 | 44.52 35 |
|
CasMVSNet(base) | | | 38.80 36 | 17.46 45 | 53.03 32 | 50.40 33 | 63.89 26 | 13.07 46 | 21.85 46 | 44.80 34 |
|
P-MVSNet | | | 38.28 37 | 29.70 33 | 44.00 40 | 39.20 42 | 50.96 44 | 29.12 31 | 30.29 36 | 41.85 37 |
|
Pnet-new- | | | 38.13 38 | 13.54 49 | 54.53 29 | 50.63 32 | 63.16 29 | 15.34 42 | 11.75 54 | 49.79 26 |
|
R-MVSNet | | | 37.16 39 | 26.49 34 | 44.27 39 | 42.37 40 | 52.02 42 | 24.06 33 | 28.93 38 | 38.41 41 |
|
AttMVS | | | 37.07 40 | 30.02 32 | 41.77 43 | 34.23 45 | 52.58 41 | 28.11 32 | 31.93 34 | 38.51 40 |
|
ANet | | | 37.00 41 | 21.13 37 | 47.57 38 | 44.38 38 | 59.75 32 | 18.53 36 | 23.72 41 | 38.59 39 |
|
Pnet_fast | | | 36.12 42 | 12.02 53 | 52.20 33 | 52.81 30 | 63.56 28 | 6.47 57 | 17.57 48 | 40.22 38 |
|
firsttry | | | 32.39 43 | 25.47 35 | 36.99 45 | 32.51 46 | 46.62 47 | 21.58 35 | 29.36 37 | 31.86 45 |
|
MVSNet_plusplus | | | 29.36 44 | 9.54 56 | 42.57 41 | 46.21 36 | 44.17 48 | 6.65 56 | 12.43 53 | 37.33 42 |
|
MVSNet | | | 27.67 45 | 18.22 44 | 33.96 46 | 26.28 50 | 43.46 50 | 13.67 45 | 22.78 44 | 32.14 44 |
|
MVE |  | 24.84 21 | 26.04 46 | 20.59 38 | 29.66 51 | 22.29 53 | 42.00 51 | 15.49 41 | 25.70 39 | 24.70 50 |
Simon Fuhrmann, Fabian Langguth, Michael Goesele: MVE - A Multi-View Reconstruction Environment. EUROGRAPHICS Workshops on Graphics and Cultural Heritage (2014) |
test_1124 | | | 25.42 47 | 7.36 59 | 37.46 44 | 31.97 47 | 49.24 46 | 8.21 54 | 6.51 59 | 31.19 46 |
|
MVSCRF | | | 24.97 48 | 13.50 50 | 32.61 49 | 26.31 49 | 43.85 49 | 12.02 48 | 14.98 50 | 27.67 47 |
|
example | | | 24.50 49 | 16.56 46 | 29.80 50 | 27.47 48 | 56.53 38 | 17.69 37 | 15.43 49 | 5.40 61 |
|
Pnet-blend | | | 23.49 50 | 9.16 57 | 33.04 47 | 38.23 43 | 35.75 54 | 3.77 61 | 14.54 51 | 25.15 48 |
|
Pnet-blend++ | | | 23.49 50 | 9.16 57 | 33.04 47 | 38.23 43 | 35.75 54 | 3.77 61 | 14.54 51 | 25.15 48 |
|
hgnet | | | 22.90 52 | 13.03 51 | 29.49 52 | 25.13 51 | 56.72 36 | 16.70 38 | 9.35 57 | 6.62 59 |
|
DPSNet | | | 22.90 52 | 13.03 51 | 29.49 52 | 25.13 51 | 56.72 36 | 16.70 38 | 9.35 57 | 6.62 59 |
|
CasMVSNet(SR_B) | | | 21.94 54 | 19.24 41 | 23.74 56 | 21.33 54 | 34.30 56 | 14.54 43 | 23.93 40 | 15.57 53 |
|
F/T MVSNet+Gipuma | | | 19.68 55 | 11.34 54 | 25.23 54 | 16.34 56 | 38.41 52 | 11.15 50 | 11.54 55 | 20.95 51 |
|
MVSNet + Gipuma | | | 19.59 56 | 11.32 55 | 25.11 55 | 16.29 58 | 38.21 53 | 11.18 49 | 11.45 56 | 20.83 52 |
|
RMVSNet | | | 15.61 57 | 22.67 36 | 10.90 59 | 13.04 60 | 14.34 60 | 12.27 47 | 33.07 31 | 5.30 62 |
|
Pnet-eth | | | 14.99 58 | 18.95 42 | 12.35 58 | 19.54 55 | 4.23 65 | 7.59 55 | 30.32 35 | 13.27 54 |
|
PMVS |  | 26.43 20 | 13.46 59 | 7.26 60 | 17.59 57 | 14.19 59 | 30.85 57 | 10.42 51 | 4.11 60 | 7.73 58 |
Y. Furukawa, J. Ponce: Accurate, dense, and robust multiview stereopsis. PAMI (2010) |
metmvs_fine | | | 12.14 60 | 14.21 48 | 10.76 60 | 8.78 61 | 13.01 62 | 10.35 52 | 18.07 47 | 10.49 55 |
|
unMVSmet | | | 7.50 61 | 3.60 61 | 10.10 61 | 6.98 62 | 18.02 59 | 3.83 60 | 3.37 61 | 5.28 63 |
|
Cas-MVS_preliminary | | | 6.17 62 | 2.14 63 | 8.86 62 | 4.86 63 | 13.44 61 | 3.08 63 | 1.20 64 | 8.27 57 |
|
test_1120 |  | | 5.30 63 | 3.59 62 | 6.45 65 | 4.63 64 | 4.96 64 | 3.98 59 | 3.19 62 | 9.76 56 |
|
CMPMVS |  | 48.56 19 | 4.84 64 | 0.02 66 | 8.06 63 | 1.22 66 | 22.96 58 | 0.03 66 | 0.00 66 | 0.00 66 |
M. Jancosek, T. Pajdla: Multi-View Reconstruction Preserving Weakly-Supported Surfaces. CVPR 2011 |
confMetMVS | | | 4.71 65 | 2.07 64 | 6.46 64 | 4.51 65 | 11.02 63 | 1.90 64 | 2.24 63 | 3.85 64 |
|
FADENet | | | 0.32 66 | 0.12 65 | 0.45 66 | 0.50 67 | 0.77 66 | 0.16 65 | 0.09 65 | 0.08 65 |
|
dnet | | | 0.00 67 | 0.00 67 | 0.00 67 | 0.00 68 | 0.00 67 | 0.00 67 | 0.00 66 | 0.00 66 |
|
UnsupFinetunedMVSNet | | | | | | 16.34 56 | | | | |
|