A nice photo of me

Daniel Worrall

d.worrall at cs.ucl.ac.uk

Erdős number: 4

Hi, I'm Daniel. I work in the Department of Computer Science at UCL in the Machine Vision Group under the supervision of Gabriel J. Brostow and Dr. Clare Wilson FRCOphth. My main interests are in scalable approximate Bayesian inference, deep learning, structured representation learning, medical imaging and optimization. I graduated from the Department of Engineering, Cambridge University in 2014 with an MEng and BA and have been at UCL since then.

My PhD is split between two major projects: I am interested in obtaining explict mathematical models of image formation and using these as useful inductive priors in convolutional neural networks e.g., harmonic networks. I am also funded by the charity Fight for Sight to automate the screening and detection of a sight-threatening, neonatal ocular disease Retinopathy of Prematurity. On the side, I am interested in useful applications of scalable Bayesian algorithms in both the vision and medical imaging domains, e.g. see our Bayesian super-resolution paper.

Google Scholar link Github link Twitter link

Looking for the UCL Computer Vision Reading Group? It's here.


Link image

Virtual Adversarial Ladder Networks For Semi-supervised Learning
Saki Shinoda, Daniel E. Worrall, Gabriel J. Brostow
Accepted to NIPS 2017 LLD Workshop

Link image

Interpretable transformations with Encoder-Decoder Networks
Daniel E. Worrall, Stephan J. Garbin, Daniyar Turmukhambetov and Gabriel J. Brostow
Accepted to ICCV 2017

Link image

Visualising The Temporal Progression Of Retinal Growth In Neonates
Daniel E. Worrall, Gabriel J. Brostow, Anna Ells, Clare M. Wilson
Accepted to World ROP Congress 2017

Link image

Bayesian Image Quality Transfer with CNNs: Exploring Uncertainty in dMRI Super-Resolution
Ryutaro Tanno, Daniel E. Worrall, Aurobrata Ghosh, Enrico Kaden, Stamatios N. Sotiropoulos, Antonio Criminisi, Daniel Alexander
Oral at MICCAI 2017 and Winner of the Young Scientist Award

Link image

Harmonic Networks: Deep Translation and Rotation Equivariance
Daniel E. Worrall, Stephan J. Garbin, Daniyar Turmukhambetov and Gabriel J. Brostow
Accepted to CVPR 2017

Link image

Automated Retinopathy of Prematurity Case Detection with Convolutional Neural Networks
Daniel E. Worrall, Clare Wilson and Gabriel J. Brostow
Deep Learning and Data Labeling for Medical Applications, Springer 2016

Link image

Automated Optic Disc Localization In The Neonatal Fundus Image
Daniel E. Worrall, Gabriel J. Brostow and Clare Wilson
Accepted to ARVO 2016

Invited Talks

Harmonic Networks: Deep Translation and Rotation Equivariance
York University, Toronto, September 2017

Video version available upon request

Reading group

I organise the UCL computer vision vision reading group. Attendence is free and open to all. We have a broad covering of topic in vision, very often branching into machine learning and computer graphics. If you have questions, or would like your email added to the mailing list, please get in touch.

Teaching and mentorship

I am a teaching assistant on Machine Vision, COMPGI14/COMPM054. I also recently co-supervised Saki Shinoda for her masters project on semi-supervised methods.

Brief Bio and CV

My Cirriculum Vitae

I grew up in Cambridge, going to school there and later reading Engineering at Sidney Sussex College, Cambridge University. As an undergraduate, I was very involved with my college engineering society and notably captained my college rowing club twice. During my masters I became interested in machine learning and computational neuroscience, choosing to write my thesis on the latter. After Cambridge I moved directly to London to begin a PhD under the supervision of Dr. Gabriel Brostow who is now Reader in Computer Vision and Computer Graphics at UCL