{"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/beyond-face-rotation-global-and-local","title":"Beyond Face Rotation: Global and Local Perception GAN for Photorealistic and Identity Preserving Frontal View Synthesis","arxiv_id":"1704.04086","date":"2017-04-13","proceeding":"ICCV 2017 10","authors":["Rui Huang","Shu Zhang","Tianyu Li","Ran He"],"abstract":"Photorealistic frontal view synthesis from a single face image has a wide\nrange of applications in the field of face recognition. Although data-driven\ndeep learning methods have been proposed to address this problem by seeking\nsolutions from ample face data, this problem is still challenging because it is\nintrinsically ill-posed. This paper proposes a Two-Pathway Generative\nAdversarial Network (TP-GAN) for photorealistic frontal view synthesis by\nsimultaneously perceiving global structures and local details. Four landmark\nlocated patch networks are proposed to attend to local textures in addition to\nthe commonly used global encoder-decoder network. Except for the novel\narchitecture, we make this ill-posed problem well constrained by introducing a\ncombination of adversarial loss, symmetry loss and identity preserving loss.\nThe combined loss function leverages both frontal face distribution and\npre-trained discriminative deep face models to guide an identity preserving\ninference of frontal views from profiles. Different from previous deep learning\nmethods that mainly rely on intermediate features for recognition, our method\ndirectly leverages the synthesized identity preserving image for downstream\ntasks like face recognition and attribution estimation. Experimental results\ndemonstrate that our method not only presents compelling perceptual results but\nalso outperforms state-of-the-art results on large pose face recognition.","url_abs":"http://arxiv.org/abs/1704.04086v2","url_pdf":"http://arxiv.org/pdf/1704.04086v2.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":"beyond-face-rotation-global-and-local","repo_url":"https://github.com/AbdielNie/GAN_Face-Rotation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"beyond-face-rotation-global-and-local","repo_url":"https://github.com/UnrealLink/TP-GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"beyond-face-rotation-global-and-local","repo_url":"https://github.com/yh-iro/Keras-TP-GAN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"face-recognition","task_name":"Face Recognition"},{"task_slug":null,"task_name":"Generative Adversarial Network"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1704.04086","atlas_url":"https://app.syntology.ai/?focus=1704.04086","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}