Self-Flow-Matching assisted Full Waveform Inversion

2026-03-12

AI Summary

This work proposes a physics-driven self-flow matching assisted full-waveform inversion (SFM-FWI) framework to address the challenges of cycle-skipping and noise sensitivity in conventional FWI when low-frequency data are missing or the initial model is inaccurate. Unlike existing diffusion-based regularization methods that rely on costly offline pretraining and are vulnerable to distribution shifts, SFM-FWI learns a transport field online, using the initial model as the dynamical starting point without assuming Gaussianity or requiring a fixed noise schedule. The method leverages FWI data residuals as a self-supervised signal, eliminating the need for offline pretraining, resolving noise-alignment ambiguities, and enabling end-to-end online optimization with arbitrary initial models. Experiments on synthetic data demonstrate that SFM-FWI significantly outperforms standard FWI and non-pretrained regularization approaches, achieving higher reconstruction accuracy, enhanced robustness to noise, and more stable convergence.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Unsupervised & Self-Supervised Learning
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Data management and stream processing for Web, mobile and wireless applications

Abstract

Full-waveform inversion (FWI) is a high-resolution seismic imaging method that estimates subsurface velocity by matching simulated and recorded waveforms. However, FWI is highly nonlinear, prone to cycle skipping, and sensitive to noise, particularly when low frequencies are missing or the initial model is poor, leading to failures under imperfect acquisition. Diffusion-regularized FWI introduces generative priors to encourage geologically realistic models, but these priors typically require costly offline pretraining and can deteriorate under distribution shift. Moreover, they assume Gaussian initialization and a fixed noise schedule, in which it is unclear how to map a deterministic FWI iterate and its starting model to a well-defined diffusion time or noise level. To address these limitations, we introduce Self-Flow-Matching assisted Full-Waveform Inversion (SFM-FWI), a physics-driven framework that eliminates the need for large-scale offline pretraining while avoiding the noise-level alignment ambiguity. SFM-FWI leverages flow matching to learn a transport field without assuming Gaussian initialization or a predefined noise schedule, so the initial model can be used directly as the starting point of the dynamics. Our approach trains a single flow network online using the governing physics and observed data. At each outer iteration, we build an interpolated model and update the flow by backpropagating the FWI data misfit, providing self-supervision without external training pairs. Experiments on challenging synthetic benchmarks show that SFM-FWI delivers more accurate reconstructions, greater noise robustness, and more stable convergence than standard FWI and pretraining-free regularization methods.

Problem

• Full-waveform inversion
• Cycle skipping
• Noise robustness
• Diffusion regularization
• Initial model bias

Innovation

• Flow Matching
• Full-Waveform Inversion
• Self-Supervised Learning
• Physics-Informed Neural Networks
• Seismic Imaging