{"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/threat-of-adversarial-attacks-on-deep","title":"Threat of Adversarial Attacks on Deep Learning in Computer Vision: A Survey","arxiv_id":"1801.00553","date":"2018-01-02","proceeding":null,"authors":["Naveed Akhtar","Ajmal Mian"],"abstract":"Deep learning is at the heart of the current rise of machine learning and\nartificial intelligence. In the field of Computer Vision, it has become the\nworkhorse for applications ranging from self-driving cars to surveillance and\nsecurity. Whereas deep neural networks have demonstrated phenomenal success\n(often beyond human capabilities) in solving complex problems, recent studies\nshow that they are vulnerable to adversarial attacks in the form of subtle\nperturbations to inputs that lead a model to predict incorrect outputs. For\nimages, such perturbations are often too small to be perceptible, yet they\ncompletely fool the deep learning models. Adversarial attacks pose a serious\nthreat to the success of deep learning in practice. This fact has lead to a\nlarge influx of contributions in this direction. This article presents the\nfirst comprehensive survey on adversarial attacks on deep learning in Computer\nVision. We review the works that design adversarial attacks, analyze the\nexistence of such attacks and propose defenses against them. To emphasize that\nadversarial attacks are possible in practical conditions, we separately review\nthe contributions that evaluate adversarial attacks in the real-world\nscenarios. Finally, we draw on the literature to provide a broader outlook of\nthe research direction.","url_abs":"http://arxiv.org/abs/1801.00553v3","url_pdf":"http://arxiv.org/pdf/1801.00553v3.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":"threat-of-adversarial-attacks-on-deep","repo_url":"https://github.com/saumya0303/Attack_on_Image_Classification_Network","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"threat-of-adversarial-attacks-on-deep","repo_url":"https://github.com/saumya0303/attack_image","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"threat-of-adversarial-attacks-on-deep","repo_url":"https://github.com/saumya0303/image-classification-network-attack","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"self-driving-cars","task_name":"Self-Driving Cars"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1801.00553","atlas_url":"https://app.syntology.ai/?focus=1801.00553","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}