Mutual information maximizing GAN inversion for real face with identity preservation |
| |
| Affiliation: | 1. Department of Information Engineering and Computer Science, Feng Chia University, Taichung 40724, Taiwan;2. Department of Computer Science, University of Warwick, Coventry CV4 7AL, UK;3. Department of Computer Science and Information Engineering, National Chin-Yi University of Technology, Taichung 41170, Taiwan;4. Department of Computer Science and Information Management, Providence University, Providence University, Taichung 433, Taiwan;1. College of Information Science and Engineering, Ningbo University, Ningbo 315211, China;2. School of Electronics and Information Engineering, Ningbo University of Technology, Ningbo 315211, China;1. Institute of Micro-Nano Optoelectronics, Shenzhen University, Shenzhen 518060, China;2. College of Mechanical and Electrical Engineering, Guangdong University of Science and Technology, Dongguan, Guangdong 523083, China;3. School of Communication and Information Engineering, Xi''an University of Posts and Telecommunications, Xi''an, Shanxi 710121, China;1. Ming Chuan University, Department of Electronic Engineering, No. 5, Deming Rd., Taoyuan City 33348, Taiwan;2. National Central University, Department of Communication Engineering, No. 300, Jhongda Rd., Taoyuan City 32001, Taiwan |
| |
| Abstract: | Recent generative adversarial networks (GANs) have yielded remarkable performance in face image synthesis. GAN inversion embeds an image into the latent space of a pretrained generator, enabling it to be used for real face manipulation. However, current inversion approaches for real faces suffer the dilemma of initialization collapse and identity loss. In this paper, we propose a hierarchical GAN inversion for real faces with identity preservation based on mutual information maximization. We first use a facial domain guaranteed initialization to avoid the initialization collapse. Furthermore, we prove that maximizing the mutual information between inverted faces and their identities is equivalent to minimizing the distance between identity features from inverted and original faces. Optimization for real face inversion with identity preservation is implemented on this mutual information-maximizing constraint. Extensive experimental results show that our approach outperforms state-of-the-art solutions for inverting and editing real faces, particularly in terms of face identity preservation. |
| |
| Keywords: | Generative adversarial network GAN inversion Mutual information maximizing Face identity preservation Face editing |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
