Augmenting neural field inputs with a clamped distance function lets a single network represent gradient discontinuities at material interfaces and creases, enabling discretization-agnostic reduced-order physics simulation.
Hugo Bertiche, Meysam Madadi, and Sergio Escalera
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.GR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Precise Gradient Discontinuities in Neural Fields for Subspace Physics
Augmenting neural field inputs with a clamped distance function lets a single network represent gradient discontinuities at material interfaces and creases, enabling discretization-agnostic reduced-order physics simulation.