2016/11/06 by Shuangfei Zhai, Yu Cheng, Zhai, Shuangfei +6 · 6 citations
Computer Science · #Digital Media Forensic Detection #FOS: Computer and information sciences #Generative Adversarial Networks and Image Synthesis #Machine Learning (cs.LG) #Music and Audio Processing #cs.LG
paper · pdf · doi:10.48550/arxiv.1611.01799
Under review at ICLR 2017
arxiv created 2016/11/06 · openalex publication_date 2016/11/06 · arxiv updated 2016/11/08 · openalex created_date 2016/11/30 · openalex updated_date 2026/07/28
In this paper, we study deep generative models for effective unsupervised learning. We propose VGAN, which works by minimizing a variational lower bound of the negative log likelihood (NLL) of an energy based model (EBM), where the model density p(x) is approximated by a variational distribution q(x) that is easy to sample from. The training of VGAN takes a two step procedure: given p(x), q(x) is updated to maximize the lower bound; p(x) is then updated one step with samples drawn from q(x) to decrease the lower bound. VGAN is inspired by the generative adversarial networks (GANs), where p(x) corresponds to the discriminator and q(x) corresponds to the generator, but with several notable differences. We hence name our model variational GANs (VGANs). VGAN provides a practical solution to training deep EBMs in high dimensional space, by eliminating the need of MCMC sampling. From this view, we are also able to identify causes to the difficulty of training GANs and propose viable solutions. \footnoteExperimental code is available at https://github.com/Shuangfei/vgan