{"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/estimating-the-success-of-unsupervised-image","title":"Estimating the Success of Unsupervised Image to Image Translation","arxiv_id":"1712.07886","date":"2017-12-21","proceeding":"ECCV 2018 9","authors":["Sagie Benaim","Tomer Galanti","Lior Wolf"],"abstract":"While in supervised learning, the validation error is an unbiased estimator\nof the generalization (test) error and complexity-based generalization bounds\nare abundant, no such bounds exist for learning a mapping in an unsupervised\nway. As a result, when training GANs and specifically when using GANs for\nlearning to map between domains in a completely unsupervised way, one is forced\nto select the hyperparameters and the stopping epoch by subjectively examining\nmultiple options. We propose a novel bound for predicting the success of\nunsupervised cross domain mapping methods, which is motivated by the recently\nproposed Simplicity Principle. The bound can be applied both in expectation,\nfor comparing hyperparameters and for selecting a stopping criterion, or per\nsample, in order to predict the success of a specific cross-domain translation.\nThe utility of the bound is demonstrated in an extensive set of experiments\nemploying multiple recent algorithms. Our code is available at\nhttps://github.com/sagiebenaim/gan_bound .","url_abs":"http://arxiv.org/abs/1712.07886v2","url_pdf":"http://arxiv.org/pdf/1712.07886v2.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":"estimating-the-success-of-unsupervised-image","repo_url":"https://github.com/sagiebenaim/gan_bound","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"generalization-bounds","task_name":"Generalization Bounds"},{"task_slug":"image-to-image-translation","task_name":"Image-to-Image Translation"},{"task_slug":"translation","task_name":"Translation"},{"task_slug":"unsupervised-image-to-image-translation","task_name":"Unsupervised Image-To-Image Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}