researcher, runner, tinkerer
x linkedin github scholar
I’m a trained researcher, finishing a PhD at the Technical University of Munich and Helmholtz Munich on lightweight models that learn from scarce data. Published at ECCV, ICCV, AAAI and MICCAI, cited over 175 times.
I’m staying in machine learning and changing what I point it at. The papers below are where I’ve been rather than where I’m going, and the next set of problems is the interesting part.
I also run and race, currently 1:45 for the half marathon and 4:08 for the full, with podium ambitions I have no intention of giving up on. The same itch shows up elsewhere: around 1700 on chess.com, and a standing habit of tinkering with things that don’t need to exist.
self-supervised and zero-shot learning, single-instance optimisation, generative models, neural implicit representations, denoising and restoration, computer vision