Improving variational autoencoder with deep feature consistent and generative adversarial training

Hou, Xianxu; Sun, Ke; Shen, Linlin; Qiu, Guoping

Abstract

We present a new method for improving the performances of variational autoencoder (VAE). In addition to enforcing the deep feature consistent principle thus ensuring the VAE output and its corresponding input images to have similar deep features, we also implement a generative adversarial training mechanism to force the VAE to output realistic and natural images. We present experimental results to show that the VAE trained with our new method outperforms state of the art in generating face images with much clearer and more natural noses, eyes, teeth, hair textures as well as reasonable backgrounds. We also show that our method can learn powerful embeddings of input face images, which can be used to achieve facial attribute manipulation. Moreover we propose a multi-view feature extraction strategy to extract effective image representations, which can be used to achieve state of the art performance in facial attribute prediction.

Citation

Hou, X., Sun, K., Shen, L., & Qiu, G. (2019). Improving variational autoencoder with deep feature consistent and generative adversarial training. Neurocomputing, 341, 183-194. https://doi.org/10.1016/j.neucom.2019.03.013

• Journal Article Type: Article
• Acceptance Date: Mar 9, 2019
• Online Publication Date: Mar 13, 2019
• Publication Date: May 14, 2019
• Deposit Date: Mar 13, 2019
• Publicly Available Date: Mar 14, 2020
• Journal: Neurocomputing
• Print ISSN: 0925-2312
• Electronic ISSN: 1872-8286
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 341
• Pages: 183-194
• DOI: https://doi.org/10.1016/j.neucom.2019.03.013
• Keywords: Image Generation; Facial Attributes; Generative model; VAE; GAN
• Public URL: https://nottingham-repository.worktribe.com/output/1635443
• Publisher URL: https://www.sciencedirect.com/science/article/pii/S0925231219303157