{"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/3d-coded-3d-correspondences-by-deep-1","title":"3D-CODED : 3D Correspondences by Deep Deformation","arxiv_id":"1806.05228","date":"2018-06-13","proceeding":null,"authors":["Thibault Groueix","Matthew Fisher","Vladimir G. Kim","Bryan C. Russell","Mathieu Aubry"],"abstract":"We present a new deep learning approach for matching deformable shapes by\nintroducing {\\it Shape Deformation Networks} which jointly encode 3D shapes and\ncorrespondences. This is achieved by factoring the surface representation into\n(i) a template, that parameterizes the surface, and (ii) a learnt global\nfeature vector that parameterizes the transformation of the template into the\ninput surface. By predicting this feature for a new shape, we implicitly\npredict correspondences between this shape and the template. We show that these\ncorrespondences can be improved by an additional step which improves the shape\nfeature by minimizing the Chamfer distance between the input and transformed\ntemplate. We demonstrate that our simple approach improves on state-of-the-art\nresults on the difficult FAUST-inter challenge, with an average correspondence\nerror of 2.88cm. We show, on the TOSCA dataset, that our method is robust to\nmany types of perturbations, and generalizes to non-human shapes. This\nrobustness allows it to perform well on real unclean, meshes from the the SCAPE\ndataset.","url_abs":"http://arxiv.org/abs/1806.05228v2","url_pdf":"http://arxiv.org/pdf/1806.05228v2.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":"3d-coded-3d-correspondences-by-deep-1","repo_url":"https://github.com/ThibaultGROUEIX/3D-CODED","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"3d-dense-shape-correspondence","task_name":"3D Dense Shape Correspondence"},{"task_slug":"3d-human-pose-estimation","task_name":"3D Human Pose Estimation"},{"task_slug":"3d-point-cloud-matching","task_name":"3D Point Cloud Matching"},{"task_slug":"3d-surface-generation","task_name":"3D Surface Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-dense-shape-correspondence-on-shrec-19","task":"3D Dense Shape Correspondence","dataset":"SHREC'19","model":"3DCODED (Trained on Surreal)","rank_in_archive_order":10,"of":11,"metrics":{"Accuracy at 1%":"2.1","Euclidean Mean Error (EME)":"8.1"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.05228","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}