{"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/pu-net-point-cloud-upsampling-network","title":"PU-Net: Point Cloud Upsampling Network","arxiv_id":"1801.06761","date":"2018-01-21","proceeding":"CVPR 2018 6","authors":["Lequan Yu","Xianzhi Li","Chi-Wing Fu","Daniel Cohen-Or","Pheng-Ann Heng"],"abstract":"Learning and analyzing 3D point clouds with deep networks is challenging due\nto the sparseness and irregularity of the data. In this paper, we present a\ndata-driven point cloud upsampling technique. The key idea is to learn\nmulti-level features per point and expand the point set via a multi-branch\nconvolution unit implicitly in feature space. The expanded feature is then\nsplit to a multitude of features, which are then reconstructed to an upsampled\npoint set. Our network is applied at a patch-level, with a joint loss function\nthat encourages the upsampled points to remain on the underlying surface with a\nuniform distribution. We conduct various experiments using synthesis and scan\ndata to evaluate our method and demonstrate its superiority over some baseline\nmethods and an optimization-based method. Results show that our upsampled\npoints have better uniformity and are located closer to the underlying\nsurfaces.","url_abs":"http://arxiv.org/abs/1801.06761v2","url_pdf":"http://arxiv.org/pdf/1801.06761v2.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":"pu-net-point-cloud-upsampling-network","repo_url":"https://github.com/yulequan/PU-Net","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"pu-net-point-cloud-upsampling-network","repo_url":"https://github.com/guochengqian/PU-GCN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"pu-net-point-cloud-upsampling-network","repo_url":"https://github.com/skoo9500/3d-pc-AE-GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"point-cloud-super-resolution","task_name":"Point Cloud Super Resolution"},{"task_slug":"point-cloud-upsampling","task_name":"point cloud upsampling"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/point-cloud-super-resolution-on-shrec15","task":"Point Cloud Super Resolution","dataset":"SHREC15","model":"PU-NET","rank_in_archive_order":3,"of":3,"metrics":{"F-measure (%)":"56.4%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1801.06761","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}