Collaborative Research: HCC: Small: Understanding Human Hair With Type 4 Simulation

Yale University, New Haven CT

Investigators

Abstract

Understanding the mechanics of human hair has wide applications in visual media and is critical to faithfully depicting virtual humans, because after the face, hair is the next characteristic that we look to when establishing a person's identity. However, almost all existing research has focused on the depiction of straight or curly hair, which spans Types 1 to 3 in the Walker hair typing system, whereas Type 4 hair, also known as afro-textured or kinky hair, has received almost no attention. This knowledge gap is unfortunate, because the naturally occurring hair type of millions of people in the United States, and a billion people worldwide, is Type 4; consequently, representing and simulating virtual Type 4 hair in digital media is difficult, because existing algorithms were not designed with its specific computational challenges in mind. More broadly, a comprehensive understanding of human hair is not possible without a complete investigation of Type 4. This project will address these deficiencies by developing new mechanical models and efficient computational methods for Type 4 hair. The fundamental mechanical behavior of a solid is determined by its strain energy, and energies are usually custom-tailored to the behavior of specific real-world solids. Leveraging the team's complementary expertise and prior experience designing robust new energies for volumetric solids, a new, anisotropic, elasto-plastic energy for Type 4 hair will be defined, and the resulting model will be validated against real-world measurements as well as extensive large-scale simulation data. Moreover, a detailed understanding of Type 4 hair will point the way toward representations that span all hair types; to bridge these types, this research will develop general eigenanalysis methods for strain energies that span the strand (1D), shell (2D), and volumetric (3D) regimes. Project outcomes will include efficient methods that are capable of generating realistic motion for human hair, in all of its diversity. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.