GAN (v11)Special Structure (v6)李宏毅深度学习备课讲稿.ppt

So many GANs …… Modifying the Optimization of GAN fGAN WGAN Least-square GAN Loss Sensitive GAN Energy-based GAN Boundary-seeking GAN Unroll GAN …… Different Structure from the Original GAN Conditional GAN Semi-supervised GAN InfoGAN BiGAN Cycle GAN Disco GAN VAE-GAN …… Conditional GAN Motivation Generator Scott Reed, Zeynep Akata, Xinchen Yan, Lajanugen Logeswaran, Bernt Schiele, Honglak Lee, “Generative Adversarial Text-to-Image Synthesis”, ICML 2016 Text Image Scott Reed,?Zeynep Akata,?Santosh Mohan,?Samuel Tenka,?Bernt Schiele,?Honglak Lee, “Learning What and Where to Draw”, NIPS 2016 Han Zhang,?Tao Xu,?Hongsheng Li,?Shaoting Zhang,?Xiaolei Huang,?Xiaogang Wang,?Dimitris Metaxas, “StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks”, arXiv prepring, 2016 Basic Idea of GAN Maximum Likelihood Estimation ? ? ? Likelihood of generating the samples ? ? Maximum Likelihood Estimation ? ? ? ? ? ? ? ? ? ? /generative-models/ ? ? ? ? It is difficult to compute the likelihood. ? Basic Idea of GAN Generator G G is a function, input z, output x Given a prior distribution Pprior(z), a probability distribution PG(x) is defined by function G Discriminator D D is a function, input x, output scalar Evaluate the “difference” between PG(x) and Pdata(x) There is a function V(G,D). ? Hard to learn by maximum likelihood Basic Idea ? ? ? ? ? ? ? ? ? ? ? ? ? ? Given G, what is the optimal D* maximizing Given x, the optimal D* maximizing ? ? ? ? ? Assume that D(x) can have any value here ? Given x, the optimal D* maximizing Find D* maximizing: ? ? ? ? ? ? ? ? ? ? a D b D 0 < < 1 ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? 2 2 ? ? ? ? ? ? ? ? ? ? ? Jensen-Shannon divergence ? ? ? ? ? ? In the end …… ? ? ? ? ? 0 < < log 2 ? Algorithm ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? Algorithm ? ? ? ? ? Decrease JS divergence(?) Decrease JS divergence(?) Algorithm ? ? ? ? Decrease JS divergence(?) ? ? ? ? ? ? smaller ? …… ? Don’t update G too much In practice … ?