MIMIR with jointly optimized TGV² regularization matches classical FMM-LSMR on Gaussian velocity models and reduces RMSE by 33-44% on layered and curved-fault models under 5% noise.
Title resolution pending
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
The paper supplies explicit hand-derived gradient formulas and a full training cycle for PINNs on a simple ODE, achieving 4.29e-4 relative L2 error against the analytic solution using only the physics loss.
citing papers explorer
-
Total Generalized Variation regularization closes the gap between neural-eld and classical methods in seismic travel-time tomography
MIMIR with jointly optimized TGV² regularization matches classical FMM-LSMR on Gaussian velocity models and reduces RMSE by 33-44% on layered and curved-fault models under 5% noise.
-
Physics-Informed Neural Networks: A Didactic Derivation of the Complete Training Cycle
The paper supplies explicit hand-derived gradient formulas and a full training cycle for PINNs on a simple ODE, achieving 4.29e-4 relative L2 error against the analytic solution using only the physics loss.