Normal and Visibility Estimation of Human Face from a Single Image

TL;DR

This paper introduces a deep learning method to estimate human face surface normals and visibility from a single image, improving detail reconstruction by decomposing light transfer functions into visibility and normal components.

Contribution

It presents a novel decomposition of light transfer functions into visibility and cosine terms, enabling better face surface normal and shading detail estimation from a single image.

Findings

Enhanced face surface normal reconstruction

Improved shading detail recovery

Better handling of visibility effects

Abstract

Recent work on the intrinsic image of humans starts to consider the visibility of incident illumination and encodes the light transfer function by spherical harmonics. In this paper, we show that such a light transfer function can be further decomposed into visibility and cosine terms related to surface normal. Such decomposition allows us to recover the surface normal in addition to visibility. We propose a deep learning-based approach with a reconstruction loss for training on real-world images. Results show that compared with previous works, the reconstruction of human face from our method better reveals the surface normal and shading details especially around regions where visibility effect is strong.