Search papers, labs, and topics across Lattice.
The paper introduces a numerical framework for exploring Blaschke-Santaló diagrams by parameterizing convex bodies using invertible neural networks based on gauge functions, ensuring convexity during shape optimization. They employ an interacting particle system that minimizes a Riesz energy functional via automatic differentiation in PyTorch to achieve uniform sampling within the diagram. The method's efficacy is demonstrated on diagrams involving geometric and PDE-type functionals for convex bodies in $\mathbb{R}^2$ and $\mathbb{R}^3$, including volume, perimeter, moment of inertia, torsional rigidity, Willmore energy, and the first two Neumann eigenvalues of the Laplacian.
Unlock new insights into shape inequalities with a neural network-based framework that efficiently explores Blaschke--Santaló diagrams.
We introduce a novel numerical framework for the exploration of Blaschke--Santaló diagrams, which are efficient tools characterizing the possible inequalities relating some given shape functionals. We introduce a parametrization of convex bodies in arbitrary dimensions using a specific invertible neural network architecture based on gauge functions, allowing an intrinsic conservation of the convexity of the sets during the shape optimization process. To achieve a uniform sampling inside the diagram, and thus a satisfying description of it, we introduce an interacting particle system that minimizes a Riesz energy functional via automatic differentiation in PyTorch. The effectiveness of the method is demonstrated on several diagrams involving both geometric and PDE-type functionals for convex bodies of $\mathbb{R}^2$ and $\mathbb{R}^3$, namely, the volume, the perimeter, the moment of inertia, the torsional rigidity, the Willmore energy, and the first two Neumann eigenvalues of the Laplacian.
