This table lists the benchmark results for the 2D Optical Flow scenario.


Method InfoAccuracy (<20px)EPE (pixel)
sorted bysort by
PWC-Netcopyleft0.74822.674
Deqing Sun, Xiaodong Yang, Ming-Yu Liu, Jan Kautz: PWC-Net: CNNs for Optical Flow Using Pyramid, Warping, and Cost Volume. CVPR 2018
FlowNet 2.0copyleft0.68527.610
Eddy Ilg, Nikolaus Mayer, Tonmoy Saikia, Margret Keuper, Alexey Dosovitskiy, Thomas Brox: FlowNet 2.0: Evolution of Optical Flow Estimation with Deep Networks. CVPR 2017

This table lists the benchmark results for the Non-rigid Reconstruction scenario.


Method InfoGeometry error (cm)Deformation error (cm)
sort bysorted by
OcclusionFusion0.3862.800
Wenbin Lin, Chengwei Zheng, Jun-Hai Yong, Feng Xu: OcclusionFusion: Occlusion-aware Motion Estimation for Real-time Dynamic 3D Reconstruction. CVPR 2022
Neural Non-rigid Tracking0.4032.872
DeepDeform0.4163.152
Aljaž Božič, Michael Zollhöfer, Christian Theobalt, Matthias Nießner: DeepDeform: Learning Non-rigid RGB-D Reconstruction with Semi-supervised Data. CVPR 2020
DynamicFusion1.0786.179
Richard Newcombe, Dieter Fox, Steve Seitz: DynamicFusion: Reconstruction and Tracking of Non-rigid Scenes in Real-Time. CVPR 2015
VolumeDeform7.48520.841
Matthias Innmann, Michael Zollhöfer, Matthias Nießner, Christian Theobalt, Marc Stamminger: VolumeDeform: Real-time Volumetric Non-rigid Reconstruction. ECCV 2016