{"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/unrestricted-facial-geometry-reconstruction","title":"Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation","arxiv_id":"1703.10131","date":"2017-03-29","proceeding":"ICCV 2017 10","authors":["Matan Sela","Elad Richardson","Ron Kimmel"],"abstract":"It has been recently shown that neural networks can recover the geometric\nstructure of a face from a single given image. A common denominator of most\nexisting face geometry reconstruction methods is the restriction of the\nsolution space to some low-dimensional subspace. While such a model\nsignificantly simplifies the reconstruction problem, it is inherently limited\nin its expressiveness. As an alternative, we propose an Image-to-Image\ntranslation network that jointly maps the input image to a depth image and a\nfacial correspondence map. This explicit pixel-based mapping can then be\nutilized to provide high quality reconstructions of diverse faces under extreme\nexpressions, using a purely geometric refinement process. In the spirit of\nrecent approaches, the network is trained only with synthetic data, and is then\nevaluated on in-the-wild facial images. Both qualitative and quantitative\nanalyses demonstrate the accuracy and the robustness of our approach.","url_abs":"http://arxiv.org/abs/1703.10131v2","url_pdf":"http://arxiv.org/pdf/1703.10131v2.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":"unrestricted-facial-geometry-reconstruction","repo_url":"https://github.com/matansel/pix2vertex","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":null},{"paper_slug":"unrestricted-facial-geometry-reconstruction","repo_url":"https://github.com/eladrich/pix2vertex.pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"3d-face-reconstruction","task_name":"3D Face Reconstruction"},{"task_slug":"image-to-image-translation","task_name":"Image-to-Image Translation"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1703.10131","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}