{"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/unpaired-photo-to-caricature-translation-on","title":"Unpaired Photo-to-Caricature Translation on Faces in the Wild","arxiv_id":"1711.10735","date":"2017-11-29","proceeding":null,"authors":["Ziqiang Zheng","Wang Chao","Zhibin Yu","Nan Wang","Haiyong Zheng","Bing Zheng"],"abstract":"Recently, image-to-image translation has been made much progress owing to the\nsuccess of conditional Generative Adversarial Networks (cGANs). And some\nunpaired methods based on cycle consistency loss such as DualGAN, CycleGAN and\nDiscoGAN are really popular. However, it's still very challenging for\ntranslation tasks with the requirement of high-level visual information\nconversion, such as photo-to-caricature translation that requires satire,\nexaggeration, lifelikeness and artistry. We present an approach for learning to\ntranslate faces in the wild from the source photo domain to the target\ncaricature domain with different styles, which can also be used for other\nhigh-level image-to-image translation tasks. In order to capture global\nstructure with local statistics while translation, we design a dual pathway\nmodel with one coarse discriminator and one fine discriminator. For generator,\nwe provide one extra perceptual loss in association with adversarial loss and\ncycle consistency loss to achieve representation learning for two different\ndomains. Also the style can be learned by the auxiliary noise input.\nExperiments on photo-to-caricature translation of faces in the wild show\nconsiderable performance gain of our proposed method over state-of-the-art\ntranslation methods as well as its potential real applications.","url_abs":"http://arxiv.org/abs/1711.10735v2","url_pdf":"http://arxiv.org/pdf/1711.10735v2.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":"unpaired-photo-to-caricature-translation-on","repo_url":"https://github.com/zhengziqiang/P2C","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"caricature","task_name":"Caricature"},{"task_slug":"image-to-image-translation","task_name":"Image-to-Image Translation"},{"task_slug":"photo-to-caricature-translation","task_name":"Photo-To-Caricature Translation"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cycle-consistency-loss","method_name":"Cycle Consistency Loss"},{"method_slug":"gan-least-squares-loss","method_name":"GAN Least Squares Loss"},{"method_slug":"instance-normalization","method_name":"Instance Normalization"},{"method_slug":"patchgan","method_name":"PatchGAN"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}