{"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-face-morphable-models-in-the-wild","title":"3D Face Morphable Models \"In-the-Wild\"","arxiv_id":"1701.05360","date":"2017-01-19","proceeding":"CVPR 2017 7","authors":["James Booth","Epameinondas Antonakos","Stylianos Ploumpis","George Trigeorgis","Yannis Panagakis","Stefanos Zafeiriou"],"abstract":"3D Morphable Models (3DMMs) are powerful statistical models of 3D facial\nshape and texture, and among the state-of-the-art methods for reconstructing\nfacial shape from single images. With the advent of new 3D sensors, many 3D\nfacial datasets have been collected containing both neutral as well as\nexpressive faces. However, all datasets are captured under controlled\nconditions. Thus, even though powerful 3D facial shape models can be learnt\nfrom such data, it is difficult to build statistical texture models that are\nsufficient to reconstruct faces captured in unconstrained conditions\n(\"in-the-wild\"). In this paper, we propose the first, to the best of our\nknowledge, \"in-the-wild\" 3DMM by combining a powerful statistical model of\nfacial shape, which describes both identity and expression, with an\n\"in-the-wild\" texture model. We show that the employment of such an\n\"in-the-wild\" texture model greatly simplifies the fitting procedure, because\nthere is no need to optimize with regards to the illumination parameters.\nFurthermore, we propose a new fast algorithm for fitting the 3DMM in arbitrary\nimages. Finally, we have captured the first 3D facial database with relatively\nunconstrained conditions and report quantitative evaluations with\nstate-of-the-art performance. Complementary qualitative reconstruction results\nare demonstrated on standard \"in-the-wild\" facial databases. An open source\nimplementation of our technique is released as part of the Menpo Project.","url_abs":"http://arxiv.org/abs/1701.05360v1","url_pdf":"http://arxiv.org/pdf/1701.05360v1.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":[],"tasks":[{"task_slug":"3d-face-reconstruction","task_name":"3D Face Reconstruction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/3d-face-reconstruction-on-florence","task":"3D Face Reconstruction","dataset":"Florence","model":"itwmm","rank_in_archive_order":6,"of":16,"metrics":{"Average 3D Error":"1.82"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1701.05360","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}