Of networks and regularization: new developments in Lagrangian Particle Tracking

This talk will introduce several recent developments in the the field of 3D particle tracking and related post-processing methods. The aspects that will be touched are: (i) Machine-learning based dectection of particle images and the implications for 3D position reconstruction. (ii) First experiences with high-performance-computing evaluations using the Shake-The-Box algorithm. (iii) Comparisons of several classes of flowfield-reconstruction approaches (Binning, Data Assimilation, Physics-Informed Neural Networks).