We propose a novel system to reconstruct a high-quality hair depth map from a single portrait photo with minimal user input. This is achieved by combining depth cues such as occlusions, silhouettes, and shading, with a novel 3D helical structural prior for hair reconstruction.

Menglei Chai, Zhejiang University
Linjie Luo, Adobe Research
Kalyan Sunkavalli, Adobe Research
Nathan Carr, Adobe Research
Sunil Hadap, Adobe Research
Kun Zhou, Zhejiang University