CaloTrilogy: Toward a Breakthrough in One-Step, End-to-End, Physics-Guided Shower Generation for Modern Calorimeters
Researchers have developed CaloTrilogy, a novel framework for high-precision calorimeter simulation in particle physics. This method combines an average velocity field integrator for fast sampling with a learned generative prior and physics-guided loss terms. The approach allows for end-to-end generation with minimal evaluation steps, achieving shower quality competitive with existing flow and diffusion models while maintaining physics fidelity. AI
IMPACT Accelerates high-precision simulation for particle physics experiments, potentially enabling faster analysis of collider data.