Explore projects

PDE inference from limited spatio-temporal data

Hexagonal spring network with pores are stretched. The resulting stress profile is analyzed. The aim is to verify the hypothesis that in the Zebrafish optic cup, ECM with more pores are under more stress than ECM with less pores.

Code accompanying "Inverse-Dirichlet Weighting Enables Reliable Training of Physics Informed Neural Networks", Maddu et al., 2021

private repo with testing jupyter notebooks

This is a python implementation of the cycle coalescence algorithm as described by Modes et al, 2016.

Exploring cellpose (deep learning) for segmentation in 2D&3D. Including setup for GPU on cluster or workstation

Implementation of 2D unet. Use as base for user projects.

This repository contains source codes and notebook demos for trainin STENCIL-NET.

Completly reimplmented Lipidxplorer, with adaptation layer to read lipidxplorer1 files

Henricks script as Python