# Learning Symmetry Consistent Deep CNNs for Face Completion

arXiv: https://arxiv.org/abs/1812.07741 · HTML: https://mlsanity.com/paper/1812.07741/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Dec 2018.

## Built on by (most cited first, 6 of 6)

- [SAFA: Structure Aware Face Animation](https://mlsanity.com/paper/2111.04928/) (2021)
- [Reference-Guided Large-Scale Face Inpainting with Identity and Texture Control](https://mlsanity.com/paper/2303.07014/) (2023)
- [Deep Face Video Inpainting via UV Mapping](https://mlsanity.com/paper/2109.00681/) (2021)
- [ShowFace: Coordinated Face Inpainting with Memory-Disentangled Refinement Networks](https://mlsanity.com/paper/2204.02824/) (2022)
- [3DFaceFill: An Analysis-By-Synthesis Approach to Face Completion](https://mlsanity.com/paper/2110.10395/) (2021)
- [Combining Generative and Geometry Priors for Wide-Angle Portrait Correction](https://mlsanity.com/paper/2410.09911/) (2024)

## Abstract (arXiv)

Deep convolutional networks (CNNs) have achieved great success in face completion to generate plausible facial structures. These methods, however, are limited in maintaining global consistency among face components and recovering fine facial details. On the other hand, reflectional symmetry is a prominent property of face image and benefits face recognition and consistency modeling, yet remaining uninvestigated in deep face completion. In this work, we leverage two kinds of symmetry-enforcing subnets to form a symmetry-consistent CNN model (i.e., SymmFCNet) for effective face completion. For missing pixels on only one of the half-faces, an illumination-reweighted warping subnet is developed to guide the warping and illumination reweighting of the other half-face. As for missing pixels on both of half-faces, we present a generative reconstruction subnet together with a perceptual symmetry loss to enforce symmetry consistency of recovered structures. The SymmFCNet is constructed by stacking generative reconstruction subnet upon illumination-reweighted warping subnet, and can be end-to-end learned from training set of unaligned face images. Experiments show that SymmFCNet can generate high quality results on images with synthetic and real occlusion, and performs favorably against state-of-the-arts.