Total Generalized Variation regularization closes the gap between neural-eld and classical methods in seismic travel-time tomography
Abstract
Travel-time tomography forces a trade-off between mesh resolution and stability in which the regularizer choice dominates what can be recovered. We introduce MIMIR, a differentiable framework that represents the 2D velocity field as a Fourier-feature neural network, replacing the grid-based slowness vector with a continuous, infinitely differentiable function. Prior neural-field tomography has staircased smooth fields under total-variation (TV) priors or oscillated near interfaces under Laplacian smoothing. We adopt second-order total generalized variation (TGV) and parametrize its auxiliary vector field as a second neural network jointly optimized with the velocity field, eliminating the inner Chambolle-Pock primal-dual loop that classically dominates TGV computation. On three synthetic benchmarks (Gaussian, horizontally layered, curved-fault inspired by OpenFWI) using cross-well acquisition, 5% travel-time noise, and five seeds, MIMIR-TGV ties a classical FMM-LSMR baseline with auto-tuned hyperparameters on the Gaussian (, paired -test) and significantly outperforms it on layered (, 44% RMSE reduction) and curved-fault (, 33% reduction). Replacing TGV with TV degrades performance on Gaussian () and layered (); curriculum-annealed TV improves Gaussian RMSE by only 5.4%, confirming that TV's staircase bias is intrinsic to the regularizer rather than a scheduling artifact. The results empirically validate the Bredies-Kunisch-Pock prediction that piecewise-affine priors are better suited to subsurface velocity recovery than piecewise-constant TV priors. We argue that the central design choice in physics-informed neural-field inversion is not the network architecture but the regularizer. The full pipeline reproduces in under one hour on consumer hardware.
Cite
@article{arxiv.2605.09960,
title = {Total Generalized Variation regularization closes the gap between neural-eld and classical methods in seismic travel-time tomography},
author = {Isao Kurosawa},
journal= {arXiv preprint arXiv:2605.09960},
year = {2026}
}
Comments
15 pages, 6 figures. Manuscript submitted to Geophysical Journal International