Understanding Trade-Offs in Continuous Neural Representations for Diffeomorphic Image Registration: A Comparative Study of Implicit Neural Representations and Neural Ordinary Differential Equations. [PDF]
Rodriguez-Sanz S +2 more
europepmc +1 more source
Selector form of Weaver's conjecture and frame sparsification
Abstract We show an extension of a probabilistic result of Marcus, Spielman, and Srivastava, which resolved the Kadison–Singer problem, to block diagonal positive semidefinite random matrices. We use this result to show several selector results, which generalize their partition counterparts.
Marcin Bownik
wiley +1 more source
Fast eikonal phase retrieval for high-throughput beamlines. [PDF]
Mirone A +5 more
europepmc +1 more source
Discrete homotopy hypothesis for n$n$‐types
Abstract We show that discrete and classical homotopy theories are equivalent after localizing at n$n$‐equivalences for any non‐negative integer n$n$. By constructing an explicit homotopy inverse to the graph nerve functor associating an n$n$‐fibrant cubical set to a graph, we are also able to give explicit computations of several previously unknown ...
Daniel Carranza, Krzysztof Kapulkin
wiley +1 more source
Rapid estimation of photon measurement density functions using a deep convolutional neural network for functional near-infrared spectroscopy. [PDF]
Zhao Y, Sun XT, Shi WD, Zhu CZ, Zhang L.
europepmc +1 more source
Abstract This study investigates the impact of phased array radar (PAR) scanning strategies on short‐term convective‐scale analyses and forecasts through an Observing System Simulation Experiment framework. The Weather Research and Forecasting model, coupled with the Data Assimilation Research Testbed ensemble adjustment Kalman filter, is used to ...
Jidong Gao, Derek R. Stratman
wiley +1 more source
On the Spectral Representation of Ordinary Self-Adjoint Differential Operators [PDF]
openaire +3 more sources
Neural Adjoint Inverse Design of Three‐Dimensional Nanophotonic Devices
A stage‐wise Fourier neural operator is trained on FDTD adjoint gradients of volumetric meta‐optics, each stage refining residuals under progressively enlarged Fourier‐mode cutoffs. Given any permittivity volume, the model reconstructs the 3D gradient field, from coarse envelope to resonance peaks, within a second, driving inverse design to 86–99% of ...
Chanik Kang +3 more
wiley +1 more source
Bayesian bilevel operator learning with low-rank adaptation for efficient uncertainty quantification of PDE inverse problems. [PDF]
Zhang RZ, Miles CE, Xie X, Lowengrub JS.
europepmc +1 more source
Electrochemical experiments reveal what happens, whereas first principles calculations explain why. Emerging computational electrochemistry is bridging these perspectives, enabling quantitative understanding of electrode/electrolyte interfaces and enabling the rational design of next‐generation energy storage materials.
Kenji Oqmhula, Ryo Maezono, Kenta Hongo
wiley +1 more source

