Presentation

· Contributors · Organizations · Search Program

Neural Geometry Fields for Meshes

DescriptionNeural geometry fields are a novel 3D surface representation which leverage advances in neural implicit functions to represent discrete triangle meshes. By discretizing a surface into quadrangular patches, a neural function can be adapted to learn high frequency surface details, ultimately enabling state-of-the-art mesh compression.

Authors

Venkataram Edavamadathil Sivaram

University of California San Diego

Tzu-Mao Li

University of California San Diego

Ravi Ramamoorthi

University of California San Diego