{"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/neural-3d-mesh-renderer","title":"Neural 3D Mesh Renderer","arxiv_id":"1711.07566","date":"2017-11-20","proceeding":"CVPR 2018 6","authors":["Hiroharu Kato","Yoshitaka Ushiku","Tatsuya Harada"],"abstract":"For modeling the 3D world behind 2D images, which 3D representation is most\nappropriate? A polygon mesh is a promising candidate for its compactness and\ngeometric properties. However, it is not straightforward to model a polygon\nmesh from 2D images using neural networks because the conversion from a mesh to\nan image, or rendering, involves a discrete operation called rasterization,\nwhich prevents back-propagation. Therefore, in this work, we propose an\napproximate gradient for rasterization that enables the integration of\nrendering into neural networks. Using this renderer, we perform single-image 3D\nmesh reconstruction with silhouette image supervision and our system\noutperforms the existing voxel-based approach. Additionally, we perform\ngradient-based 3D mesh editing operations, such as 2D-to-3D style transfer and\n3D DeepDream, with 2D supervision for the first time. These applications\ndemonstrate the potential of the integration of a mesh renderer into neural\nnetworks and the effectiveness of our proposed renderer.","url_abs":"http://arxiv.org/abs/1711.07566v1","url_pdf":"http://arxiv.org/pdf/1711.07566v1.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":"neural-3d-mesh-renderer","repo_url":"https://github.com/laughtervv/tf_neural_renderer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"neural-3d-mesh-renderer","repo_url":"https://github.com/mattiagaggi/3D-Pose-Estimator---MSc_project","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"neural-3d-mesh-renderer","repo_url":"https://github.com/monniert/unicorn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"3d-object-reconstruction","task_name":"3D Object Reconstruction"},{"task_slug":"style-transfer","task_name":"Style Transfer"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-object-reconstruction-on-data3dr2n2","task":"3D Object Reconstruction","dataset":"Data3D−R2N2","model":"N3MR","rank_in_archive_order":15,"of":15,"metrics":{"Avg F1":"33.80"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1711.07566","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}