{"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/loss-sensitive-generative-adversarial","title":"Loss-Sensitive Generative Adversarial Networks on Lipschitz Densities","arxiv_id":"1701.06264","date":"2017-01-23","proceeding":null,"authors":["Guo-Jun Qi"],"abstract":"In this paper, we present the Lipschitz regularization theory and algorithms\nfor a novel Loss-Sensitive Generative Adversarial Network (LS-GAN).\nSpecifically, it trains a loss function to distinguish between real and fake\nsamples by designated margins, while learning a generator alternately to\nproduce realistic samples by minimizing their losses. The LS-GAN further\nregularizes its loss function with a Lipschitz regularity condition on the\ndensity of real data, yielding a regularized model that can better generalize\nto produce new data from a reasonable number of training examples than the\nclassic GAN. We will further present a Generalized LS-GAN (GLS-GAN) and show it\ncontains a large family of regularized GAN models, including both LS-GAN and\nWasserstein GAN, as its special cases. Compared with the other GAN models, we\nwill conduct experiments to show both LS-GAN and GLS-GAN exhibit competitive\nability in generating new images in terms of the Minimum Reconstruction Error\n(MRE) assessed on a separate test set. We further extend the LS-GAN to a\nconditional form for supervised and semi-supervised learning problems, and\ndemonstrate its outstanding performance on image classification tasks.","url_abs":"http://arxiv.org/abs/1701.06264v6","url_pdf":"http://arxiv.org/pdf/1701.06264v6.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":"loss-sensitive-generative-adversarial","repo_url":"https://github.com/guojunq/lsgan","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok"}}],"tasks":[{"task_slug":null,"task_name":"Generative Adversarial Network"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-classification-on-cifar-10","task":"Image Classification","dataset":"CIFAR-10","model":"CLS-GAN","rank_in_archive_order":190,"of":265,"metrics":{"Percentage correct":"91.7"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-svhn","task":"Image Classification","dataset":"SVHN","model":"CLS-GAN","rank_in_archive_order":43,"of":62,"metrics":{"Percentage error":"5.98"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1701.06264","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}