Deep Single-Image Portrait Relighting

By Hao Zhou, Sunil Hadap, Kalyan Sunkavalli, David W. Jacobs. 2019

255 citations recorded by the source corpus.

Conventional physically-based methods for relighting portrait images need to solve an inverse rendering problem, estimating face geometry, reflectance and lighting. However, the inaccurate estimation of face components can cause strong artifacts in relighting, leading to unsatisfactory results. In this work, we apply a physically-based portrait relighting method to generate a large scale, high quality, “in the wild” portrait relighting dataset (DPR). A deep Convolutional Neural Network (CNN) is then trained using this dataset to generate a relit portrait image by using a source image and a target lighting as input. The training procedure regularizes the generated results, removing the artifacts caused by physically-based relighting methods. A GAN loss is further applied to improve the quality of the relit portrait image. Our trained network can relight portrait images with resolutions as high as 1024 × 1024. We evaluate the proposed method on the proposed DPR datset, Flickr portrait dataset and Multi-PIE dataset both qualitatively and quantitatively. Our experiments demonstrate that the proposed method achieves state-of-the-art results. Please refer to https://zhhoper.github.io/dpr.html for dataset and code.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition, Generative Models

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

In Generative Models

Citations in the corpus

GAN Inversion: A Survey (2021, background): “, gaze redirection [6], [179], relighting [180], [181], [182], and continuous view control [183].”
Portrait shadow manipulation (2020, methodology): “Our method can also be used as a “preprocessing” step for image modification algorithms such as portrait relighting (Sun et al. 2019; Zhou et al. 2019), which modify or replace the illumination of the input image.”
Disentangled and Controllable Face Image Generation via 3D Imitative-Contrastive Learning (2020, background): “GANs have seen heavy use in face image manipulation [34, 19, 49, 44, 36, 45, 53].”
PIE (2020, methodology): “We can edit colored illumination while Zhou et al. [2019] can only edit monochrome light.”
TediGAN: Text-Guided Diverse Image Generation and Manipulation (2020, unknown): “Those conditioned on control vectors are typically limited to a small range of applications such as gaze redi-rection [13] or relighting [44, 59].”