Wispy to Voluminous: Prior-free Multi-view Capture of Strand-level Facial Hair

A new multi-view capture pipeline can turn facial hair into editable strand assets instead of baking it into face geometry. For teams shipping avatars, that matters because beards, brows, lashes, and mustaches are often where volumetric shortcuts fall apart. The method aims to preserve sparse, wispy structure while still supporting animation, grooming, and simulation.