{"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/soft-rasterizer-a-differentiable-renderer-for","title":"Soft Rasterizer: A Differentiable Renderer for Image-based 3D Reasoning","arxiv_id":"1904.01786","date":"2019-04-03","proceeding":"ICCV 2019 10","authors":["Shichen Liu","Tianye Li","Weikai Chen","Hao Li"],"abstract":"Rendering bridges the gap between 2D vision and 3D scenes by simulating the\nphysical process of image formation. By inverting such renderer, one can think\nof a learning approach to infer 3D information from 2D images. However,\nstandard graphics renderers involve a fundamental discretization step called\nrasterization, which prevents the rendering process to be differentiable, hence\nable to be learned. Unlike the state-of-the-art differentiable renderers, which\nonly approximate the rendering gradient in the back propagation, we propose a\ntruly differentiable rendering framework that is able to (1) directly render\ncolorized mesh using differentiable functions and (2) back-propagate efficient\nsupervision signals to mesh vertices and their attributes from various forms of\nimage representations, including silhouette, shading and color images. The key\nto our framework is a novel formulation that views rendering as an aggregation\nfunction that fuses the probabilistic contributions of all mesh triangles with\nrespect to the rendered pixels. Such formulation enables our framework to flow\ngradients to the occluded and far-range vertices, which cannot be achieved by\nthe previous state-of-the-arts. We show that by using the proposed renderer,\none can achieve significant improvement in 3D unsupervised single-view\nreconstruction both qualitatively and quantitatively. Experiments also\ndemonstrate that our approach is able to handle the challenging tasks in\nimage-based shape fitting, which remain nontrivial to existing differentiable\nrenderers.","url_abs":"http://arxiv.org/abs/1904.01786v1","url_pdf":"http://arxiv.org/pdf/1904.01786v1.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":"soft-rasterizer-a-differentiable-renderer-for","repo_url":"https://github.com/ShichenLiu/SoftRas","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"soft-rasterizer-a-differentiable-renderer-for","repo_url":"https://github.com/kach/softraxterizer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"jax","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"3d-object-reconstruction","task_name":"3D Object Reconstruction"},{"task_slug":"single-view-3d-reconstruction","task_name":"Single-View 3D Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-object-reconstruction-on-shapenet","task":"3D Object Reconstruction","dataset":"ShapeNet","model":"SoftRas (full)","rank_in_archive_order":1,"of":1,"metrics":{"3DIoU":"0.6464"},"uses_additional_data":false},{"leaderboard":"/sota/single-view-3d-reconstruction-on-shapenet","task":"Single-View 3D Reconstruction","dataset":"ShapeNet","model":"SoftRas (full)","rank_in_archive_order":3,"of":7,"metrics":{"3DIoU":"0.6464"},"uses_additional_data":false},{"leaderboard":"/sota/single-view-3d-reconstruction-on-shapenet","task":"Single-View 3D Reconstruction","dataset":"ShapeNet","model":"NMR [19]","rank_in_archive_order":5,"of":7,"metrics":{"3DIoU":"0.6015"},"uses_additional_data":false},{"leaderboard":"/sota/single-view-3d-reconstruction-on-shapenet","task":"Single-View 3D Reconstruction","dataset":"ShapeNet","model":"voxel [47]","rank_in_archive_order":6,"of":7,"metrics":{"3DIoU":"0.5736"},"uses_additional_data":false},{"leaderboard":"/sota/single-view-3d-reconstruction-on-shapenet","task":"Single-View 3D Reconstruction","dataset":"ShapeNet","model":"retrieval [47]","rank_in_archive_order":7,"of":7,"metrics":{"3DIoU":"0.4766"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1904.01786","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1904.01786"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ShichenLiu/SoftRas","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kach/softraxterizer","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":1,"unverified":10},"by_repo_kind":{"official":{"samples":9,"ran":1,"repositories":1},"listed":{"samples":2,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"42d983df1501fbb5","entry":"face_vertices","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/face_vertices.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/face_vertices.py","link_basis":"plan_row","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"42d983df1501fbb5"}},{"code_sha256_prefix":"7ffb516ee103321a","entry":"ambient_lighting","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/ambient_lighting.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/ambient_lighting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7ffb516ee103321a"}},{"code_sha256_prefix":"74684204a6a3b4c6","entry":"directional_lighting","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/directional_lighting.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/directional_lighting.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"74684204a6a3b4c6"}},{"code_sha256_prefix":"01749edd95bc6e58","entry":"get_pixel","repo":"kach/softraxterizer","repo_kind":"listed","path":"softraxterizer.py","file_url":"https://github.com/kach/softraxterizer/blob/HEAD/softraxterizer.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"01749edd95bc6e58"}},{"code_sha256_prefix":"8497e22d3575f176","entry":"get_points_from_angles","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/get_points_from_angles.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/get_points_from_angles.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8497e22d3575f176"}},{"code_sha256_prefix":"158489df7e42d7e0","entry":"load_mtl","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/load_obj.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/load_obj.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"158489df7e42d7e0"}},{"code_sha256_prefix":"e13da960399c788d","entry":"look","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/look.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/look.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e13da960399c788d"}},{"code_sha256_prefix":"d7e7c5fad58018a4","entry":"look_at","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/functional/look_at.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/functional/look_at.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d7e7c5fad58018a4"}},{"code_sha256_prefix":"e26d5f3fd5f9dd7c","entry":"orthogonal","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/transform.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/transform.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e26d5f3fd5f9dd7c"}},{"code_sha256_prefix":"be8ae6f86624b34a","entry":"perspective","repo":"ShichenLiu/SoftRas","repo_kind":"official","path":"soft_renderer/transform.py","file_url":"https://github.com/ShichenLiu/SoftRas/blob/HEAD/soft_renderer/transform.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"be8ae6f86624b34a"}},{"code_sha256_prefix":"b8b5518bfb5539c8","entry":"softras","repo":"kach/softraxterizer","repo_kind":"listed","path":"softraxterizer.py","file_url":"https://github.com/kach/softraxterizer/blob/HEAD/softraxterizer.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b8b5518bfb5539c8"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}