ECO combines a strongly constrained equilibrium-compatibility U-Net with low-resolution supervision and a soft constitutive law to super-resolve micromechanical fields without high-resolution training simulations.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Equilibrium Conserving Neural Operators for Super-Resolution Learning
ECO combines a strongly constrained equilibrium-compatibility U-Net with low-resolution supervision and a soft constitutive law to super-resolve micromechanical fields without high-resolution training simulations.