Saved in:
Bibliographic Details
Main Authors: Peng, Jishen, Jiang, Enze, Ma, Zheng, Yan, Xiongbin
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2603.16393
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910056154398720
author Peng, Jishen
Jiang, Enze
Ma, Zheng
Yan, Xiongbin
author_facet Peng, Jishen
Jiang, Enze
Ma, Zheng
Yan, Xiongbin
contents We develop a robust physics-guided diffusion framework for full-waveform inversion that combines a score-based generative prior with likelihood guidance computed through wave-equation simulations. We adopt a transport-based data-consistency potential (Wasserstein-2), incorporating wavefield enhancement via bounded weighting and observation-dependent normalization, thereby improving robustness to amplitude imbalance and time/phase misalignment. On the inference side, we introduce a preconditioned guided reverse-diffusion scheme that adapts the guidance strength and spatial scaling throughout the reverse-time dynamics, yielding a more stable and effective data-consistency guidance step than standard diffusion posterior sampling (DPS). Numerical experiments on OpenFWI datasets demonstrate improved reconstruction quality over deterministic optimization baselines and standard DPS under comparable computational budgets.
format Preprint
id arxiv_https___arxiv_org_abs_2603_16393
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Robust Physics-Guided Diffusion for Full-Waveform Inversion
Peng, Jishen
Jiang, Enze
Ma, Zheng
Yan, Xiongbin
Numerical Analysis
Artificial Intelligence
We develop a robust physics-guided diffusion framework for full-waveform inversion that combines a score-based generative prior with likelihood guidance computed through wave-equation simulations. We adopt a transport-based data-consistency potential (Wasserstein-2), incorporating wavefield enhancement via bounded weighting and observation-dependent normalization, thereby improving robustness to amplitude imbalance and time/phase misalignment. On the inference side, we introduce a preconditioned guided reverse-diffusion scheme that adapts the guidance strength and spatial scaling throughout the reverse-time dynamics, yielding a more stable and effective data-consistency guidance step than standard diffusion posterior sampling (DPS). Numerical experiments on OpenFWI datasets demonstrate improved reconstruction quality over deterministic optimization baselines and standard DPS under comparable computational budgets.
title Robust Physics-Guided Diffusion for Full-Waveform Inversion
topic Numerical Analysis
Artificial Intelligence
url https://arxiv.org/abs/2603.16393