{"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/on-detecting-adversarial-perturbations","title":"On Detecting Adversarial Perturbations","arxiv_id":"1702.04267","date":"2017-02-14","proceeding":null,"authors":["Jan Hendrik Metzen","Tim Genewein","Volker Fischer","Bastian Bischoff"],"abstract":"Machine learning and deep learning in particular has advanced tremendously on\nperceptual tasks in recent years. However, it remains vulnerable against\nadversarial perturbations of the input that have been crafted specifically to\nfool the system while being quasi-imperceptible to a human. In this work, we\npropose to augment deep neural networks with a small \"detector\" subnetwork\nwhich is trained on the binary classification task of distinguishing genuine\ndata from data containing adversarial perturbations. Our method is orthogonal\nto prior work on addressing adversarial perturbations, which has mostly focused\non making the classification network itself more robust. We show empirically\nthat adversarial perturbations can be detected surprisingly well even though\nthey are quasi-imperceptible to humans. Moreover, while the detectors have been\ntrained to detect only a specific adversary, they generalize to similar and\nweaker adversaries. In addition, we propose an adversarial attack that fools\nboth the classifier and the detector and a novel training procedure for the\ndetector that counteracts this attack.","url_abs":"http://arxiv.org/abs/1702.04267v2","url_pdf":"http://arxiv.org/pdf/1702.04267v2.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":"on-detecting-adversarial-perturbations","repo_url":"https://github.com/yaoshiang/MobileNetV2-CIFAR-Cleverhans","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"adversarial-attack","task_name":"Adversarial Attack"},{"task_slug":"binary-classification","task_name":"Binary Classification"},{"task_slug":"classification","task_name":"General Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1702.04267","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}