{"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/generative-modeling-using-the-sliced","title":"Generative Modeling using the Sliced Wasserstein Distance","arxiv_id":"1803.11188","date":"2018-03-29","proceeding":"CVPR 2018 6","authors":["Ishan Deshpande","Ziyu Zhang","Alexander Schwing"],"abstract":"Generative Adversarial Nets (GANs) are very successful at modeling\ndistributions from given samples, even in the high-dimensional case. However,\ntheir formulation is also known to be hard to optimize and often not stable.\nWhile this is particularly true for early GAN formulations, there has been\nsignificant empirically motivated and theoretically founded progress to improve\nstability, for instance, by using the Wasserstein distance rather than the\nJenson-Shannon divergence. Here, we consider an alternative formulation for\ngenerative modeling based on random projections which, in its simplest form,\nresults in a single objective rather than a saddle-point formulation. By\naugmenting this approach with a discriminator we improve its accuracy. We found\nour approach to be significantly more stable compared to even the improved\nWasserstein GAN. Further, unlike the traditional GAN loss, the loss formulated\nin our method is a good measure of the actual distance between the\ndistributions and, for the first time for GAN training, we are able to show\nestimates for the same.","url_abs":"http://arxiv.org/abs/1803.11188v1","url_pdf":"http://arxiv.org/pdf/1803.11188v1.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":"generative-modeling-using-the-sliced","repo_url":"https://github.com/ishansd/swg","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.11188","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}