Solving Inverse Physics Problems with Score Matching
This work addresses inverse problems in physics for researchers and engineers, offering a novel approach that improves upon existing methods.
The authors tackled inverse physics problems by developing a method that combines an approximate inverse physics simulator with a learned correction function, achieving excellent accuracy and temporal stability while enabling posterior sampling of solutions.
We propose to solve inverse problems involving the temporal evolution of physics systems by leveraging recent advances from diffusion models. Our method moves the system's current state backward in time step by step by combining an approximate inverse physics simulator and a learned correction function. A central insight of our work is that training the learned correction with a single-step loss is equivalent to a score matching objective, while recursively predicting longer parts of the trajectory during training relates to maximum likelihood training of a corresponding probability flow. We highlight the advantages of our algorithm compared to standard denoising score matching and implicit score matching, as well as fully learned baselines for a wide range of inverse physics problems. The resulting inverse solver has excellent accuracy and temporal stability and, in contrast to other learned inverse solvers, allows for sampling the posterior of the solutions.