{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/dual-generator-generative-adversarial","title":"Dual Generator Generative Adversarial Networks for Multi-Domain Image-to-Image Translation","arxiv_id":"1901.04604","date":"2019-01-14","proceeding":null,"authors":["Hao Tang","Dan Xu","Wei Wang","Yan Yan","Nicu Sebe"],"abstract":"State-of-the-art methods for image-to-image translation with Generative\nAdversarial Networks (GANs) can learn a mapping from one domain to another\ndomain using unpaired image data. However, these methods require the training\nof one specific model for every pair of image domains, which limits the\nscalability in dealing with more than two image domains. In addition, the\ntraining stage of these methods has the common problem of model collapse that\ndegrades the quality of the generated images. To tackle these issues, we\npropose a Dual Generator Generative Adversarial Network (G$^2$GAN), which is a\nrobust and scalable approach allowing to perform unpaired image-to-image\ntranslation for multiple domains using only dual generators within a single\nmodel. Moreover, we explore different optimization losses for better training\nof G$^2$GAN, and thus make unpaired image-to-image translation with higher\nconsistency and better stability. Extensive experiments on six publicly\navailable datasets with different scenarios, i.e., architectural buildings,\nseasons, landscape and human faces, demonstrate that the proposed G$^2$GAN\nachieves superior model capacity and better generation performance comparing\nwith existing image-to-image translation GAN models.","url_abs":"http://arxiv.org/abs/1901.04604v1","url_pdf":"http://arxiv.org/pdf/1901.04604v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"dual-generator-generative-adversarial","repo_url":"https://github.com/Ha0Tang/AsymmetricGAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":null,"task_name":"Generative Adversarial Network"},{"task_slug":"image-to-image-translation","task_name":"Image-to-Image Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1901.04604","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}