{"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/pluralistic-image-completion","title":"Pluralistic Image Completion","arxiv_id":"1903.04227","date":"2019-03-11","proceeding":"CVPR 2019 6","authors":["Chuanxia Zheng","Tat-Jen Cham","Jianfei Cai"],"abstract":"Most image completion methods produce only one result for each masked input,\nalthough there may be many reasonable possibilities. In this paper, we present\nan approach for \\textbf{pluralistic image completion} -- the task of generating\nmultiple and diverse plausible solutions for image completion. A major\nchallenge faced by learning-based approaches is that usually only one ground\ntruth training instance per label. As such, sampling from conditional VAEs\nstill leads to minimal diversity. To overcome this, we propose a novel and\nprobabilistically principled framework with two parallel paths. One is a\nreconstructive path that utilizes the only one given ground truth to get prior\ndistribution of missing parts and rebuild the original image from this\ndistribution. The other is a generative path for which the conditional prior is\ncoupled to the distribution obtained in the reconstructive path. Both are\nsupported by GANs. We also introduce a new short+long term attention layer that\nexploits distant relations among decoder and encoder features, improving\nappearance consistency. When tested on datasets with buildings (Paris), faces\n(CelebA-HQ), and natural images (ImageNet), our method not only generated\nhigher-quality completion results, but also with multiple and diverse plausible\noutputs.","url_abs":"http://arxiv.org/abs/1903.04227v2","url_pdf":"http://arxiv.org/pdf/1903.04227v2.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":"pluralistic-image-completion","repo_url":"https://github.com/lyndonzheng/Pluralistic-Inpainting","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"diversity","task_name":"Diversity"},{"task_slug":"image-inpainting","task_name":"Image Inpainting"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.04227","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}