82 citations · 98 across the 7 of their papers we have counts for
8 papers
Content-Adaptive Downsampling in Convolutional Neural Networks
Robin Hesse, Simone Schaub-Meyer, Stefan Roth
Many convolutional neural networks (CNNs) rely on progressive downsampling of their feature maps to increase the network's receptive field and decrease computational cost. However,…
Fast Axiomatic Attribution for Neural Networks
Robin Hesse, Simone Schaub-Meyer, Stefan Roth
Mitigating the dependence on spurious correlations present in the training dataset is a quickly emerging and important topic of deep learning. Recent approaches include priors on t…
Global solutions for semilinear rough partial differential equations
Robert Hesse, Alexandra Neamtu
We construct global-in-time solutions for semilinear parabolic rough partial differential equations. We work on a scale of Banach spaces tailored to the controlled rough path appro…
Global solutions and random dynamical systems for rough evolution equations
Robert Hesse, Alexandra Neamtu
We consider infinite-dimensional parabolic rough evolution equations. Using regularizing properties of analytic semigroups we prove global-in-time existence of solutions and invest…
Local mild solutions for rough stochastic partial differential equations
Robert Hesse, Alexandra Neamtu
We investigate mild solutions for stochastic evolution equations driven by a fractional Brownian motion (fBm) with Hurst parameter H in (1/3, 1/2] in infinite-dimensional Banach sp…
Proximal Heterogeneous Block Input-Output Method and application to Blind Ptychographic Diffraction Imaging
Robert Hesse, D. Russell Luke, Shoham Sabach +1
We propose a general alternating minimization algorithm for nonconvex optimization problems with separable structure and nonconvex coupling between blocks of variables. To fix our…
Alternating Projections and Douglas-Rachford for Sparse Affine Feasibility
Robert Hesse, D. Russell Luke, Patrick Neumann
The problem of finding a vector with the fewest nonzero elements that satisfies an underdetermined system of linear equations is an NP-complete problem that is typically solved num…
Nonconvex notions of regularity and convergence of fundamental algorithms for feasibility problems
Robert Hesse, D. Russell Luke
We consider projection algorithms for solving (nonconvex) feasibility problems in Euclidean spaces. Of special interest are the Method of Alternating Projections (MAP) and the Doug…