{"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/generative-adversarial-perturbations","title":"Generative Adversarial Perturbations","arxiv_id":"1712.02328","date":"2017-12-06","proceeding":"CVPR 2018 6","authors":["Omid Poursaeed","Isay Katsman","Bicheng Gao","Serge Belongie"],"abstract":"In this paper, we propose novel generative models for creating adversarial\nexamples, slightly perturbed images resembling natural images but maliciously\ncrafted to fool pre-trained models. We present trainable deep neural networks\nfor transforming images to adversarial perturbations. Our proposed models can\nproduce image-agnostic and image-dependent perturbations for both targeted and\nnon-targeted attacks. We also demonstrate that similar architectures can\nachieve impressive results in fooling classification and semantic segmentation\nmodels, obviating the need for hand-crafting attack methods for each task.\nUsing extensive experiments on challenging high-resolution datasets such as\nImageNet and Cityscapes, we show that our perturbations achieve high fooling\nrates with small perturbation norms. Moreover, our attacks are considerably\nfaster than current iterative methods at inference time.","url_abs":"http://arxiv.org/abs/1712.02328v3","url_pdf":"http://arxiv.org/pdf/1712.02328v3.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":"generative-adversarial-perturbations","repo_url":"https://github.com/OmidPoursaeed/Generative_Adversarial_Perturbations","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1712.02328","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}