Alexander Chudasama
Thesis title: Self-supervised Learning Reconstruction Methods for Low-dose Biological Electron Microscopy
Franklin supervisors: Amirafshar Moshtaghpour, Mark Basham, Angus Kirkland
University: University of Edinburgh
University supervisor: Mike Davies
Electron microscopy image data typically has an extremely low SNR (appearing as grainy static at first glance). We aim to denoise these images using self-supervised deep learning techniques, allowing us to resolve biological structures present.
Alex graduated from Durham University with an MSci in Natural Sciences, but he calls himself a mathematician. His masters thesis was on studying the numerical continuation of reaction diffusion systems. Prior to this he developed a prototype of a rapid parallelised solver for partial differential equations, now known as VisualPDE! Following two years in industry, he applied to the Franklin and is looking forward to the interdisciplinary work the institute has to offer. Besides mathematics he can often be found playing the cello and bouldering/falling!