{"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/weakly-supervised-3d-reconstruction-with","title":"Weakly supervised 3D Reconstruction with Adversarial Constraint","arxiv_id":"1705.10904","date":"2017-05-31","proceeding":null,"authors":["JunYoung Gwak","Christopher B. Choy","Animesh Garg","Manmohan Chandraker","Silvio Savarese"],"abstract":"Supervised 3D reconstruction has witnessed a significant progress through the\nuse of deep neural networks. However, this increase in performance requires\nlarge scale annotations of 2D/3D data. In this paper, we explore inexpensive 2D\nsupervision as an alternative for expensive 3D CAD annotation. Specifically, we\nuse foreground masks as weak supervision through a raytrace pooling layer that\nenables perspective projection and backpropagation. Additionally, since the 3D\nreconstruction from masks is an ill posed problem, we propose to constrain the\n3D reconstruction to the manifold of unlabeled realistic 3D shapes that match\nmask observations. We demonstrate that learning a log-barrier solution to this\nconstrained optimization problem resembles the GAN objective, enabling the use\nof existing tools for training GANs. We evaluate and analyze the manifold\nconstrained reconstruction on various datasets for single and multi-view\nreconstruction of both synthetic and real images.","url_abs":"http://arxiv.org/abs/1705.10904v2","url_pdf":"http://arxiv.org/pdf/1705.10904v2.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":"weakly-supervised-3d-reconstruction-with","repo_url":"https://github.com/chrischoy/3D-R2N2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"weakly-supervised-3d-reconstruction-with","repo_url":"https://github.com/pranavbajoria93/3D_Reconstruction_3DR2N2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"3d-reconstruction","task_name":"3D Reconstruction"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1705.10904","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}