Linda-Sophie Schneider

Researcher · Pattern Recognition Lab, FAU Erlangen-Nürnberg
UnLost — ERC Synergy Grant

• Ph.D. candidate, Pattern Recognition Lab, FAU Erlangen-Nürnberg
• Differentiable CT reconstruction & CT trajectory optimization
• Co-developer of DiffCT-MLX (CUDA/PyTorch & Metal/MLX)
• Part of UnLost: reconstruction & ink detection for the Herculaneum papyri

I studied mathematics in Erlangen and am currently pursuing my Ph.D. in computer science at the Friedrich-Alexander-Universität Erlangen-Nürnberg, at the Pattern Recognition Lab. My research sits at the intersection of machine learning, differentiable programming, and tomographic reconstruction, specifically the question of how physical models of X-ray imaging can be embedded directly into learnable pipelines. I co-develop DiffCT-MLX, an open-source dual-backend toolkit for differentiable CT that supports both CUDA/PyTorch and Metal/MLX. I am particularly drawn to problems where a well-understood physical forward model and a data-driven method have to be reconciled rather than played off against each other.

In UnLost I work on the reconstruction of the carbonised scrolls and palimpsests, and on building an end-to-end pipeline that runs from segmentation to ink detection.

My fiancé does cryptography!

Email: linda-sophie.schneider@fau.de