ETH Zurich and EPFL Unveil PLATON, an AI Camera That Reconstructs Particle Tracks From a Handful of Photons

Researchers at ETH Zurich and EPFL have demonstrated PLATON, a prototype particle-detection system that fuses a single scintillator block, a light-field camera and a Transformer neural network to reconstruct 3D particle trajectories from as few as five detected photons. The work, published in *Nature Communications*, points to a radically simpler way to build the massive detectors used in physics and could eventually sharpen medical imaging as well.