Face de-occlusion

Face de-occlusion using 3d morphable model and generative adversarial network

Feature

A novel method is proposed to restore de-occluded face images based on the use of 3DMM and generative adversarial network. Experiments shows the advantages of this method on challenging facial de-occlusion, 3D face reconstruction and face attibute editing.

Face de-occlusion on synthetic images

(a) Occluded-images (b) De-occluded images (c) Real images

Face de-occlusion on real images

Dataset and code

If you are interested in this work, you can download:

Dataset [baidu drive] [google drive] Experimental Result [baidu drive] password: 2ub4

Code [coming soon]

If you use this dataset, please cite to the papers:

[1] Xiaowei Yuan and In Kyu Park. Face de-occlusion using 3d morphable model and generative adversarial network. In ICCV, 2019

[2] Tal Hassner, Shai Harel, Eran Paz, and Roee Enbar. Effective face frontalization in unconstrained images. In CVPR, 2015

[3] Xiangyu Zhu, Zhen Lei, Junjie Yan, Dong Yi, and Stan Z. Li. High-fidelity pose and expression normalization for face recognition in the wild. In CVPR, 2015.