Unsupervised Cross-Domain Image Generation

Yaniv Taigman, Adam Polyak, Lior Wolf

Nov 04, 2016 (modified: Mar 12, 2017) ICLR 2017 conference submission readers: everyone
  • Abstract: We study the problem of transferring a sample in one domain to an analog sample in another domain. Given two related domains, S and T, we would like to learn a generative function G that maps an input sample from S to the domain T, such that the output of a given representation function f, which accepts inputs in either domains, would remain unchanged. Other than f, the training data is unsupervised and consist of a set of samples from each domain, without any mapping between them. The Domain Transfer Network (DTN) we present employs a compound loss function that includes a multiclass GAN loss, an f preserving component, and a regularizing component that encourages G to map samples from T to themselves. We apply our method to visual domains including digits and face images and demonstrate its ability to generate convincing novel images of previously unseen entities, while preserving their identity.
  • Conflicts: fb.com, tau.ac.il
  • Keywords: Computer vision, Deep learning, Unsupervised Learning, Transfer Learning