About me
Contents
Research
I’m a PhD research student at the Institute for Imaging, Data and Communications at the University of Edinburgh, supervised by Prof Mike Davies.
I’m developing unsupervised ML methodologies for solving inverse imaging problems. These problems appear in many critical imaging scenarios: accelerating medical imaging, higher resolution Earth observation, electron microscopy etc. We rarely have ground truth in the real world - how do we know what a black hole is supposed to look like before ever imaging a black hole - motivating the need for provable unsupervised techniques.
Get in touch if you are interested in computer vision, inverse problems, unsupervised deep learning, geometric deep learning, compressed sensing, medical imaging or Earth observation.
Check out my CV and Google Scholar.
Publications
- A. Wang, M. Davies, “Fully unsupervised dynamic MRI reconstruction via diffeo-temporal equivariance”, on arXiv, Oct 2024. Blog.
- A. Wang, M. Davies, “Perspective-Equivariant Imaging: an Unsupervised Framework for Multispectral Pansharpening”, ECCV Workshop on Traditional Computer Vision in the Age of Deep Learning, 2024. Blog.
- J. Walsh, O. Kesa, A. Wang et al., “Near Real-Time Social Distance Estimation in London”, The Computer Journal, 2023. Winner of OUP Wilkes Award 2024. Press. Blog.
- P. Houdouin, A. Wang et al., “Robust Classification with Flexible Discriminant Analysis in Heterogeneous Data”, IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022. Blog.
Workshops & Conferences
- 10-06-2024 | “Perspective-Equivariant Imaging: an Unsupervised Framework for Pansharpening”, Maths4DL Geometric Deep Learning workshop lightning talk & poster, University of Cambridge.
Software
I also am a core contributor to the DeepInverse, a PyTorch-based library for solving imaging inverse problems with deep learning. Some recent contributions include:
- Generative adversarial networks for image reconstruction
- Image transformations for equivariant imaging
- Invertible image transformations
- Self-supervised algorithms including Artifact2Artifact and Phase2Phase
- Dynamic Magnetic Resonance Imaging (MRI) physics, 3D and multi-coil MRI
Data Scientist
Check out my project portfolio.
🏞️🏞️🏞️
Activism
I’m an activist and campaigner for better access to nature and the outdoors for everyone. See below for previous writing and speaking.
I founded ESEA Outdoors UK (about), a grassroots community to advocate for better representation, diversity and inclusion in the outdoors.
I’m a Trustee of National Trails UK, where my interest is in enabling underrepresented and marginalised people to access nature.
I’m also interested in the data of UK land access rights.
To learn more about these topics, check out our list of Resources and Substack.
Writing
- 02-12-2024 | Substack | Against inclusivity-washing in the outdoors
- 15-10-2024 | Substack | Access to nature - but for whom?
- 18-06-2024 | BMC Summit magazine | Diversity in the Mountains at the Great Lakeland 3 Day (free to read)
- 21-03-2024 | BMC | East and Southeast Asian people go outdoors too
- 18-09-2023 | Earth.org | Why Improving UK Land Access Rights Is Important For A More Sustainable Outdoors
- 16-02-2023 | Earth.org | Comparing Urban Environmental Sustainability Indicators In Europe
Speaker
- 23-11-2024 | Kendal Mountain Festival | The Human Powered Session - presented by Patagonia with Eben Myrddin Muse and Carlos Casas.
- 14-07-2024 | Love Trails Festival | Land Access: the need for a true Right to Roam with Eben Myrddin Muse (BMC), Jon Moses (Right to Roam) and Amy-Jane Beer
- 12-07-2024 | Love Trails Festival | The Outsiders Project: remarkable people doing remarkable things with Phil Young and Trina Dawkins.
- 07-07-2024 | Timber Festival | Using community to tackle racism in the outdoors: an ESEA perspective
- 05-05-2024 | SCARPA Great Lakeland 3 Day | Diversity & inclusion in the outdoors
- 28-11-2020 | Kendal Mountain Literature Festival | Open Mountain - Space and Isolation
Get in touch
Get in touch via email or LinkedIn.