User profiles for Daniel Worrall

Daniel Worrall

University of Amsterdam
Verified email at deepmind.com
Cited by 2840

Harmonic networks: Deep translation and rotation equivariance

DE Worrall, SJ Garbin… - Proceedings of the …, 2017 - openaccess.thecvf.com
Translating or rotating an input image should not affect the results of many computer vision
tasks. Convolutional neural networks (CNNs) are already translation equivariant: input image …

[PDF][PDF] Se (3)-transformers: 3d roto-translation equivariant attention networks

F Fuchs, D Worrall, V Fischer… - Advances in neural …, 2020 - proceedings.neurips.cc
We introduce the SE (3)-Transformer, a variant of the self-attention module for 3D point-clouds,
which is equivariant under continuous 3D roto-translations. Equivariance is important to …

Deep scale-spaces: Equivariance over scale

D Worrall, M Welling - Advances in Neural Information …, 2019 - proceedings.neurips.cc
We introduce deep scale-spaces, a generalization of convolutional neural networks, exploiting
the scale symmetry structure of conventional image recognition tasks. Put plainly, the …

[HTML][HTML] SARS-CoV-2 viral load is associated with increased disease severity and mortality

…, H Corry, C Wong, A Rosenthal, D Worrall… - Nature …, 2020 - nature.com
The relationship between SARS-CoV-2 viral load and risk of disease progression remains
largely undefined in coronavirus disease 2019 (COVID-19). Here, we quantify SARS-CoV-2 …

Cubenet: Equivariance to 3d rotation and translation

D Worrall, G Brostow - Proceedings of the European …, 2018 - openaccess.thecvf.com
3D Convolutional Neural Networks are sensitive to transformations applied to their input.
This is a problem because a voxelized version of a 3D object, and its rotated clone, will look …

Message passing neural PDE solvers

J Brandstetter, D Worrall, M Welling - arXiv preprint arXiv:2202.03376, 2022 - arxiv.org
The numerical solution of partial differential equations (PDEs) is difficult, having led to a
century of research so far. Recently, there have been pushes to build neural--numerical hybrid …

Profound Treg perturbations correlate with COVID-19 severity

…, A Piechocka-Trocha, DP Worrall… - Proceedings of the …, 2021 - National Acad Sciences
The hallmark of severe COVID-19 is an uncontrolled inflammatory response, resulting from
poorly understood immunological dysfunction. We hypothesized that perturbations in FoxP3 …

Learning likelihoods with conditional normalizing flows

C Winkler, D Worrall, E Hoogeboom, M Welling - 2019 - openreview.net
Normalizing Flows (NFs) are able to model complicated distributions p(y) with strong inter-dimensional
correlations and high multimodality by transforming a simple base density p(z) …

Mdp homomorphic networks: Group symmetries in reinforcement learning

E Van der Pol, D Worrall, H van Hoof… - Advances in …, 2020 - proceedings.neurips.cc
This paper introduces MDP homomorphic networks for deep reinforcement learning. MDP
homomorphic networks are neural networks that are equivariant under symmetries in the joint …

Bayesian image quality transfer with CNNs: exploring uncertainty in dMRI super-resolution

R Tanno, DE Worrall, A Ghosh, E Kaden… - … Image Computing and …, 2017 - Springer
In this work, we investigate the value of uncertainty modelling in 3D super-resolution with
convolutional neural networks (CNNs). Deep learning has shown success in a plethora of …