{"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/extracting-triangular-3d-models-materials-and","title":"Extracting Triangular 3D Models, Materials, and Lighting From Images","arxiv_id":"2111.12503","date":"2021-11-24","proceeding":"CVPR 2022 1","authors":["Jacob Munkberg","Jon Hasselgren","Tianchang Shen","Jun Gao","Wenzheng Chen","Alex Evans","Thomas Müller","Sanja Fidler"],"abstract":"We present an efficient method for joint optimization of topology, materials and lighting from multi-view image observations. Unlike recent multi-view reconstruction approaches, which typically produce entangled 3D representations encoded in neural networks, we output triangle meshes with spatially-varying materials and environment lighting that can be deployed in any traditional graphics engine unmodified. We leverage recent work in differentiable rendering, coordinate-based networks to compactly represent volumetric texturing, alongside differentiable marching tetrahedrons to enable gradient-based optimization directly on the surface mesh. Finally, we introduce a differentiable formulation of the split sum approximation of environment lighting to efficiently recover all-frequency lighting. Experiments show our extracted models used in advanced scene editing, material decomposition, and high quality view interpolation, all running at interactive rates in triangle-based renderers (rasterizers and path tracers). Project website: https://nvlabs.github.io/nvdiffrec/ .","url_abs":"https://arxiv.org/abs/2111.12503v5","url_pdf":"https://arxiv.org/pdf/2111.12503v5.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":"extracting-triangular-3d-models-materials-and","repo_url":"https://github.com/NVlabs/nvdiffrec","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"extracting-triangular-3d-models-materials-and","repo_url":"https://github.com/nvlabs/tiny-cuda-nn","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"depth-prediction","task_name":"Depth Prediction"},{"task_slug":"image-relighting","task_name":"Image Relighting"},{"task_slug":"inverse-rendering","task_name":"Inverse Rendering"},{"task_slug":"surface-normals-estimation","task_name":"Surface Normals Estimation"},{"task_slug":"surface-reconstruction","task_name":"Surface Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-relighting-on-stanford-orb","task":"Image Relighting","dataset":"Stanford-ORB","model":"NVDiffRec","rank_in_archive_order":6,"of":7,"metrics":{"HDR-PSNR":"22.91","LPIPS":"0.039","SSIM":"0.963"},"uses_additional_data":false},{"leaderboard":"/sota/inverse-rendering-on-stanford-orb","task":"Inverse Rendering","dataset":"Stanford-ORB","model":"NVDiffRec","rank_in_archive_order":6,"of":7,"metrics":{"HDR-PSNR":"22.91"},"uses_additional_data":false},{"leaderboard":"/sota/surface-normals-estimation-on-stanford-orb","task":"Surface Normals Estimation","dataset":"Stanford-ORB","model":"NVDiffRec","rank_in_archive_order":2,"of":7,"metrics":{"Cosine Distance":"0.06"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2111.12503","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}