{"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/geometry-consistent-generative-adversarial","title":"Geometry-Consistent Generative Adversarial Networks for One-Sided Unsupervised Domain Mapping","arxiv_id":"1809.05852","date":"2018-09-16","proceeding":"CVPR 2019 6","authors":["Huan Fu","Mingming Gong","Chaohui Wang","Kayhan Batmanghelich","Kun Zhang","DaCheng Tao"],"abstract":"Unsupervised domain mapping aims to learn a function to translate domain X to\nY by a function GXY in the absence of paired examples. Finding the optimal GXY\nwithout paired data is an ill-posed problem, so appropriate constraints are\nrequired to obtain reasonable solutions. One of the most prominent constraints\nis cycle consistency, which enforces the translated image by GXY to be\ntranslated back to the input image by an inverse mapping GYX. While cycle\nconsistency requires the simultaneous training of GXY and GY X, recent studies\nhave shown that one-sided domain mapping can be achieved by preserving pairwise\ndistances between images. Although cycle consistency and distance preservation\nsuccessfully constrain the solution space, they overlook the special properties\nthat simple geometric transformations do not change the semantic structure of\nimages. Based on this special property, we develop a geometry-consistent\ngenerative adversarial network (GcGAN), which enables one-sided unsupervised\ndomain mapping. GcGAN takes the original image and its counterpart image\ntransformed by a predefined geometric transformation as inputs and generates\ntwo images in the new domain coupled with the corresponding\ngeometry-consistency constraint. The geometry-consistency constraint reduces\nthe space of possible solutions while keep the correct solutions in the search\nspace. Quantitative and qualitative comparisons with the baseline (GAN alone)\nand the state-of-the-art methods including CycleGAN and DistanceGAN demonstrate\nthe effectiveness of our method.","url_abs":"http://arxiv.org/abs/1809.05852v2","url_pdf":"http://arxiv.org/pdf/1809.05852v2.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":"geometry-consistent-generative-adversarial","repo_url":"https://github.com/hufu6371/gcgan","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":null,"task_name":"Generative Adversarial Network"}],"methods":[{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cycle-consistency-loss","method_name":"Cycle Consistency Loss"},{"method_slug":"gan-least-squares-loss","method_name":"GAN Least Squares Loss"},{"method_slug":"instance-normalization","method_name":"Instance Normalization"},{"method_slug":"patchgan","method_name":"PatchGAN"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.05852","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}