AdjointFlow: Accelerating Scientific Inverse Problems with Physics-Based Transport
cs.LG
Submitted: 2026-11-10
Updated: 2026-11-10
Comments: 51 pages, 17 figures, including appendices
Code: https://github.com/maguerrap/Inverse_Scattering
License: http://creativecommons.org/licenses/by/4.0/
The gist: Inverse problems are ubiquitous across scientific applications ranging from seismic imaging to astrophysics to nondestructive testing…
Terminology
Abstract
Inverse problems are ubiquitous across scientific applications ranging from seismic imaging to astrophysics to nondestructive testing, yet classical solvers often struggle with nonconvexity and repeated evaluations of expensive physical models. Diffusion models can amortize this computation offline, but largely data-driven conditional models underutilize known physics and suffer from ill-conditioned transport maps when the target distribution is concentrated. Plug-and-play (PnP) methods explicitly leverage the physical model to guide a learned prior, improving data consistency but reintroducing costly iterative computations. We instead propose AdjointFlow, a framework built on stochastic interpolants that injects the known physics into the generative transport itself. We design a data-dependent, adjoint-based initialization to reduce transport distance to the target, modify the transport velocity to remain regular as the endpoint distributions concentrate, and parameterize the network to build the initialization directly into the learned velocity. Together, these enable accurate reconstruction at substantially lower cost than generative models starting from Gaussian noise. Experiments on inverse scattering, seismic imaging, and black hole imaging show that AdjointFlow consistently produces more accurate reconstructions than conditional diffusion models while being up to five orders of magnitude faster at inference than PnP methods. Optional inference-time refinement that queries the physical model can further improve data consistency, matching or exceeding the accuracy of PnP methods without retraining.
Sources
- Self-Flow-Matching assisted Full Waveform Inversion
- Gaussian-Channel Regularity and Distribution Learning with Flow Matching
- On the Bernstein-Von Mises Theorem for High Dimensional Nonlinear Bayesian Inverse Problems
- Adam: A Method for Stochastic Optimization
- fastMRI: An Open Dataset and Benchmarks for Accelerated MRI
- Probabilistic Inversion with Flow Matching
- Post-FWI Injection of Learned Priors Using a Flow Matching Model
- Conditional flow matching for physics-constrained inverse problems with finite training data
- Projection Embedded Diffusion Bridge for CT Reconstruction from Incomplete Data
- Variational Optimality of F\"ollmer Processes in Generative Diffusions
- ScoreField: Neural Inverse Scattering with Score-Based Generative Priors
- Plug-and-Play Latent Diffusion for Electromagnetic Inverse Scattering with Application to Brain Imaging
- Diffusion Model-Based Posterior Sampling in Full Waveform Inversion
- Optimal Transport Tools (OTT): A JAX Toolbox for all things Wasserstein
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis