Improving Full Waveform Inversion in Large Model Era

TL;DR

This paper demonstrates that large, scaled models trained on simple synthetic seismic data can generalize effectively to complex, real-world geological structures in full waveform inversion, achieving state-of-the-art results.

Contribution

It introduces a scaling strategy across model capacity, data diversity, and training methods that enables generalization of data-driven FWI models to realistic geological benchmarks.

Findings

Achieves state-of-the-art performance on OpenFWI.

Significantly narrows the generalization gap in FWI.

Successfully infers complex structures absent from training data.

Abstract

Full Waveform Inversion (FWI) is a highly nonlinear and ill-posed problem that aims to recover subsurface velocity maps from surface-recorded seismic waveforms data. Existing data-driven FWI typically uses small models, as available datasets have limited volume, geological diversity, and spatial extent, leading to substantial concerns about overfitting. Although they perform well on synthetic datasets, current methods fail to generalize to more realistic geological structures. In this work, we show that a model trained entirely on simulated and relatively simple data can generalize remarkably well to challenging and unseen geological benchmarks. We provide a working recipe that tames a billion-parameter model for FWI through coordinated scaling across three axes: model capacity, data diversity, and training strategy. Our model achieves state-of-the-art performance on OpenFWI and significantly narrows the generalization gap in data-driven FWI. Across six challenging geophysical benchmarks, including Marmousi, 2D SEG/EAGE Salt and Overthrust, 2004 BP, Sigsbee, and SEAM Phase I, it infers complex structures absent from the training set and delivers significant performance improvements (SSIM from 0.5844 to 0.7669). Overall, our results demonstrate that with an appropriate scaling strategy, large models trained on simple synthetic data can achieve substantial generalization to more complex and realistic geological structures.