Catching up with missing particles

The implementation of particle-tracking techniques with deep neural networks is a promising way to determine particle motion within complex flow structures. A graph neural network-enhanced method enables accurate particle tracking by significantly reducing the number of lost trajectories.