4 papers
De-homogenization using Convolutional Neural Networks
Martin O. Elingaard, Niels Aage, J. Andreas Bærentzen +1
This paper presents a deep learning-based de-homogenization method for structural compliance minimization. By using a convolutional neural network to parameterize the mapping from…
Shape from Projections via Differentiable Forward Projector for Computed Tomography
Jakeoung Koo, Anders B. Dahl, J. Andreas Bærentzen +3
In computed tomography, the reconstruction is typically obtained on a voxel grid. In this work, however, we propose a mesh-based reconstruction method. For tomographic problems, 3D…
De-homogenization of optimal multi-scale 3D topologies
Jeroen Groen, Florian Stutz, Niels Aage +2
This paper presents a highly efficient method to obtain high-resolution, near-optimal 3D topologies optimized for minimum compliance on a standard PC. Using an implicit geometry de…
Geometry of turbulent dissipation and the Navier-Stokes regularity problem
Janet Rafner, Zoran Grujić, Christian Bach +6
The question of whether a singularity can form in an initially regular flow, described by the 3D incompressible Navier-Stokes (NS) equations, is a fundamental problem in mathematic…