Visual Information Processing and Learning
Visual Information Processing and Learning


News / Events

One paper of VIPL is accepted by IEEE TIP

Date of publication:2019-05-05      Number of clicks: 100

Congratulations! One paper of VIPL “AttGAN: Facial Attribute Editing by Only Changing What You Want” is accepted by IEEE Transactions on Image Processing (IEEE TIP).


  Facial attribute editing aims to edit a face image by manipulating single or multiple attributes of interest, e.g., hair color, expression, mustache, and age. Two main objectives of facial attribute editing are 1) accurately editing the attributes, 2) well preserving the other details. Most of the existing methods cannot simultaneous satisfy these two objectives. Besides, most existing methods use distinct models for different attributes, which is difficult for real deployment.


  In this work, we propose an AttGAN method, which appropriately uses the attribute classification constraint, the reconstruction learning, and the adversarial learning based on the encoder-decoder architecture, achieving promising performance for multiple facial attribute editing with a single model. The main contribution of this work includes: 1) properly considering the relation between the attributes and the face latent representation under the principle of just satisfying the correct editing objective; 2) incorporating the attribute classification constraint, the reconstruction learning, and the adversarial learning into a unified framework for high-quality facial attribute editing; 3) promising results of multiple facial attribute editing using a single model.



Zhenliang He, Wangmeng Zuo, Meina Kan, Shiguang Shan and Xilin Chen, "AttGAN: Facial Attribute Editing by Only Changing What You Want," IEEE Transactions on Image Processing (TIP), 2019. (Accepted, arXiv:1711.10678).



Visual Information Processing and Learning
  • Address :No.6 Kexueyuan South Road
  • Zhongguancun,Haidian District
  • Beijing,China
  • Postcode :100190
  • Tel : (8610)62600514
  • Email:yi.cheng@vipl.ict.ac.cn
  • Valse

  • Big Lecture of DL

Copyright @ Visual Information Processing and Learning 京ICP备05002829号 京公网安备1101080060