Julius B. Kirkegaard

The paper introduces a discrete inverse rendering framework using integer programming to solve multiple biological image analysis tasks by jointly optimizing discrete decisions across time and space.

DNAsight automates AFM image analysis of chromatin organization, revealing protein-specific structural signatures and nucleosome spacing distributions.

This study uses differentiable programming on a vertex model to explore how cell-level forces can be tuned to achieve predefined tissue shapes.

This paper introduces a diffusion model-based method for instance segmentation that spontaneously breaks symmetry, allowing overlapping labels in cell instance segmentation.

The paper identifies three structural features of chemical reaction networks that predict out-of-steady-state relaxation dynamics, applicable to both artificial and biological systems.