This workshop is targeted at early career researchers working in geometry, machine learning, and biomedical engineering. It aims to bring together researchers from the three disciplines to foster exchange and collaboration.
The first day features two tutorials: one applied tutorial on geometric machine learning implementations in Python, and one theoretical tutorial on differential geometry and comparing different notions of convolutions. The first tutorial consists of a pre-recorded part, and one Q&A and discussion session during the workshop on the tutorial topic. Attendees should watch the pre-recorded part if they plan to participate in this section. Apart from this, there will be five research talks on applications of geometry to machine learning and biomedical engineering.
There will be a poster session on the first day during which participants are invited to give a 5 minutes poster presentation about their own research. The poster content need not be original research, but can be. The posters are mainly a help to get everyone to know each other.
The seminar will be held via Zoom, and the link will be distributed to registered participants before the event. To register, please fill in the registration form here: Link. Slots for the poster session are allocated on a first-come-first-served basis. (If you previously registered via email you need not fill in the form)
Pre-recorded content:
- |
Matthias Fey (TU Dortmund University): Tutorial Link: https://www.ucl.ac.uk/junior-geometry/Geo-ML-Health/Fey-video.html |
Schedule 24 September 2020 (times in British Summer Time BST):
14.30-16.00 BST |
"What do you do?" poster presentations |
16.00-16.20 BST |
Break |
16.20-16.50 BST |
Matthias Fey (TU Dortmund University): Q&A and discussion for tutorial, please watch the pre-recorded content if you plan to attend this segment (video) |
16.50-17.10 BST |
Break |
17.10-18.10 BST |
Momchil Konstantinov (Eigen Technologies): Tutorial "Convolutions Galore" (slides) (video) |
18.10-18.30 BST |
Break |
18.30-19.30 BST |
Yi Zhou (Adobe): Fully Convolutional Mesh Autoencoder Using Spatially Efficient Varying Kernels (abstract) (slides) (video) |
Schedule 25 September 2020 (times in British Summer Time BST):
13.30-14.30 BST |
Angelica I. Aviles-Rivero (University of Cambridge): Learning to Classify Large-Scale Medical Data with Minimal Supervision on Graphs (abstract) |
14.30-14.50 BST |
Break |
14.50-15.50 BST |
Sofia Ira Ktena (Deepmind): Graph Neural Networks in Connectomics and beyond (abstract) (video) |
15.50-16.10 BST |
Break |
16.10-17.10 BST |
Jean Feydy (Imperial College London): Geometric data analysis, beyond convolutions (abstract) (slides) (video) |
17.10-17.30 BST |
Break |
17.30-18.30 BST |
Guadalupe Gonzalez-Pigorini (Imperial College London): Applications of graph neural networks in computational biology (abstract) |
Organisers: Simone Foti (Wellcome/EPSRC Centre for Interventional and Surgical Sciences, UCL and Centre For Medical Image Computing UCL, simone.foti.18@ucl.ac.uk), Daniel Platt (Imperial College London, London School of Number Theory and Geometry, d.platt17@imperial.ac.uk)