Skip to main navigation Skip to search Skip to main content

Facial expression synthesis by u-net conditional generative adversarial networks

  • Beihang University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

High-level manipulation of facial expressions in images such as expression synthesis is challenging because facial expression changes are highly non-linear, and vary depending on the facial appearance. Identity of the person should also be well preserved in the synthesized face. In this paper, we propose a novel U-Net Conditioned Generative Adversarial Network (UC-GAN) for facial expression generation. U-Net helps retain the property of the input face, including the identity information and facial details. We also propose an identity preserving loss, which further improves the performance of our model. Both qualitative and quantitative experiments are conducted on the Oulu-CASIA and KDEF datasets, and the results show that our method can generate faces with natural and realistic expressions while preserve the identity information. Comparison with the state-of-the-art approaches also demonstrates the competency of our method.

Original languageEnglish
Title of host publicationICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval
PublisherAssociation for Computing Machinery, Inc
Pages283-290
Number of pages8
ISBN (Print)9781450350464
DOIs
StatePublished - 5 Jun 2018
Event8th ACM International Conference on Multimedia Retrieval, ICMR 2018 - Yokohama, Japan
Duration: 11 Jun 201814 Jun 2018

Publication series

NameICMR 2018 - Proceedings of the 2018 ACM International Conference on Multimedia Retrieval

Conference

Conference8th ACM International Conference on Multimedia Retrieval, ICMR 2018
Country/TerritoryJapan
CityYokohama
Period11/06/1814/06/18

Keywords

  • Facial expression synthesis
  • Generative adversarial network (gan)
  • Identity preserving

Fingerprint

Dive into the research topics of 'Facial expression synthesis by u-net conditional generative adversarial networks'. Together they form a unique fingerprint.

Cite this