Event

Proposes cage-based deformation structural affinity regularization algorithm

1 reports · 1 sources

Aviraj Newatia et al. propose Structural Affinity in arXiv, an explicit regularization method for cage-based deformation (CBD): it precomputes an affinity matrix using mean value coordinates, binds inner control cages to coarser outer control cages, couples control handle motion, enables smooth propagation of displacements, reduces cage self-intersection and mesh distortion, and adds only linear overhead during deformation. The authors report that in three tasks—robot control-morphology co-design, source-to-target neural deformation transfer, and pneumatic wing shape optimization—the method produces more regular deformations and achieves consistent improvements on three diagnostics: cotangent Laplacian, cage distortion, and surface roughness, while reducing reliance on hand-designed structural loss terms.