AEGAN: Generating imperceptible face synthesis via autoencoder-based generative adversarial network

Aolin Che, Jing Hua Yang, Cai Guo, Hong Ning Dai, Haoran Xie*, Ping Li*

*Corresponding author for this work

Research output: Contribution to journalJournal articlepeer-review


Face recognition (FR) systems based on convolutional neural networks have shown excellent performance in human face inference. However, some malicious users may exploit such powerful systems to identify others' face images disclosed by victims' social network accounts, consequently obtaining private information. To address this emerging issue, synthesizing face protection images with visual and protective effects is essential. However, existing face protection methods encounter three critical problems: poor visual effect, limited protective effect, and trade-off between visual and protective effects. To address these challenges, we propose a novel face protection approach in this article. Specifically, we design a generative adversarial network (GAN) framework with an autoencoder (AEGAN) as the generator to synthesize the protection images. It is worth noting that we introduce an interpolation upsampling module in the decoder in order to let the synthesized protection images evade recognition by powerful convolution-based FR systems. Furthermore, we introduce an attention module with a perceptual loss in AEGAN to enhance the visual effects of synthesized images by AEGAN. Extensive experiments have shown that AEGAN not only can maintain the comfortable visual quality of synthesized images but also prevent the recognition of commercial FR systems, including Baidu and iKLYTEK.

Original languageEnglish
Article numbere2160
JournalComputer Animation and Virtual Worlds
Issue number3-4
Publication statusPublished - 9 May 2023

Scopus Subject Areas

  • Software
  • Computer Graphics and Computer-Aided Design

User-Defined Keywords

  • autoencoder
  • face image protection
  • face recognition
  • multi-head attention


Dive into the research topics of 'AEGAN: Generating imperceptible face synthesis via autoencoder-based generative adversarial network'. Together they form a unique fingerprint.

Cite this