Multi-view face synthesis via progressive face flow

Publication Type

Journal Article

Publication Date

1-2021

Abstract

Existing GAN-based multi-view face synthesis methods rely heavily on "creating" faces, and thus they struggle in reproducing the faithful facial texture and fail to preserve identity when undergoing a large angle rotation. In this paper, we combat this problem by dividing the challenging large-angle face synthesis into a series of easy small-angle rotations, and each of them is guided by a face flow to maintain faithful facial details. In particular, we propose a Face Flow-guided Generative Adversarial Network (FFlowGAN) that is specifically trained for small-angle synthesis. The proposed network consists of two modules, a face flow module that aims to compute a dense correspondence between the input and target faces. It provides strong guidance to the second module, face synthesis module, for emphasizing salient facial texture. We apply FFlowGAN multiple times to progressively synthesize different views, and therefore facial features can be propagated to the target view from the very beginning. All these multiple executions are cascaded and trained end-to-end with a unified back-propagation, and thus we ensure each intermediate step contributes to the final result. Extensive experiments demonstrate the proposed divide-and-conquer strategy is effective, and our method outperforms the state-of-the-art on four benchmark datasets qualitatively and quantitatively.

Keywords

Faces;Face recognition;Generative adversarial networks;Image reconstruction;Facial features;Deep learning;Three-dimensional displays;Multi-view face synthesis;pose-invariant face recognition;face reconstruction

Discipline

Information Security

Research Areas

Information Systems and Management

Publication

IEEE Transactions on Image Processing

Volume

30

First Page

6024

Last Page

6035

ISSN

1057-7149

Identifier

10.1109/TIP.2021.3090658

Publisher

Institute of Electrical and Electronics Engineers

Additional URL

https://doi.org/10.1109/TIP.2021.3090658

This document is currently not available here.

Share

COinS