{"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/photorealistic-facial-texture-inference-using","title":"Photorealistic Facial Texture Inference Using Deep Neural Networks","arxiv_id":"1612.00523","date":"2016-12-02","proceeding":"CVPR 2017 7","authors":["Shunsuke Saito","Lingyu Wei","Liwen Hu","Koki Nagano","Hao Li"],"abstract":"We present a data-driven inference method that can synthesize a\nphotorealistic texture map of a complete 3D face model given a partial 2D view\nof a person in the wild. After an initial estimation of shape and low-frequency\nalbedo, we compute a high-frequency partial texture map, without the shading\ncomponent, of the visible face area. To extract the fine appearance details\nfrom this incomplete input, we introduce a multi-scale detail analysis\ntechnique based on mid-layer feature correlations extracted from a deep\nconvolutional neural network. We demonstrate that fitting a convex combination\nof feature correlations from a high-resolution face database can yield a\nsemantically plausible facial detail description of the entire face. A complete\nand photorealistic texture map can then be synthesized by iteratively\noptimizing for the reconstructed feature correlations. Using these\nhigh-resolution textures and a commercial rendering framework, we can produce\nhigh-fidelity 3D renderings that are visually comparable to those obtained with\nstate-of-the-art multi-view face capture systems. We demonstrate successful\nface reconstructions from a wide range of low resolution input images,\nincluding those of historical figures. In addition to extensive evaluations, we\nvalidate the realism of our results using a crowdsourced user study.","url_abs":"http://arxiv.org/abs/1612.00523v1","url_pdf":"http://arxiv.org/pdf/1612.00523v1.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":"photorealistic-facial-texture-inference-using","repo_url":"https://github.com/rjaisw12/3DFaceFitting","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"face-model","task_name":"Face Model"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1612.00523","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}