Neural Particle Automata: Learning Self-Organizing Particle Dynamics
This work addresses the challenge of learning self-organizing dynamics in particle-based systems for applications in computer graphics and simulation, representing an incremental advancement over existing methods.
The authors tackled the problem of extending Neural Cellular Automata to dynamic particle systems by introducing Neural Particle Automata, which model cells as particles with continuous positions and internal states updated by a learnable neural rule, achieving scalability and enabling new behaviors in tasks like morphogenesis and point-cloud classification.
We introduce Neural Particle Automata (NPA), a Lagrangian generalization of Neural Cellular Automata (NCA) from static lattices to dynamic particle systems. Unlike classical Eulerian NCA where cells are pinned to pixels or voxels, NPA model each cell as a particle with a continuous position and internal state, both updated by a shared, learnable neural rule. This particle-based formulation yields clear individuation of cells, allows heterogeneous dynamics, and concentrates computation only on regions where activity is present. At the same time, particle systems pose challenges: neighborhoods are dynamic, and a naive implementation of local interactions scale quadratically with the number of particles. We address these challenges by replacing grid-based neighborhood perception with differentiable Smoothed Particle Hydrodynamics (SPH) operators backed by memory-efficient, CUDA-accelerated kernels, enabling scalable end-to-end training. Across tasks including morphogenesis, point-cloud classification, and particle-based texture synthesis, we show that NPA retain key NCA behaviors such as robustness and self-regeneration, while enabling new behaviors specific to particle systems. Together, these results position NPA as a compact neural model for learning self-organizing particle dynamics.