{"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/decoupling-direction-and-norm-for-efficient","title":"Decoupling Direction and Norm for Efficient Gradient-Based L2 Adversarial Attacks and Defenses","arxiv_id":"1811.09600","date":"2018-11-23","proceeding":null,"authors":["Jérôme Rony","Luiz G. Hafemann","Luiz S. Oliveira","Ismail Ben Ayed","Robert Sabourin","Eric Granger"],"abstract":"Research on adversarial examples in computer vision tasks has shown that\nsmall, often imperceptible changes to an image can induce misclassification,\nwhich has security implications for a wide range of image processing systems.\nConsidering $L_2$ norm distortions, the Carlini and Wagner attack is presently\nthe most effective white-box attack in the literature. However, this method is\nslow since it performs a line-search for one of the optimization terms, and\noften requires thousands of iterations. In this paper, an efficient approach is\nproposed to generate gradient-based attacks that induce misclassifications with\nlow $L_2$ norm, by decoupling the direction and the norm of the adversarial\nperturbation that is added to the image. Experiments conducted on the MNIST,\nCIFAR-10 and ImageNet datasets indicate that our attack achieves comparable\nresults to the state-of-the-art (in terms of $L_2$ norm) with considerably\nfewer iterations (as few as 100 iterations), which opens the possibility of\nusing these attacks for adversarial training. Models trained with our attack\nachieve state-of-the-art robustness against white-box gradient-based $L_2$\nattacks on the MNIST and CIFAR-10 datasets, outperforming the Madry defense\nwhen the attacks are limited to a maximum norm.","url_abs":"http://arxiv.org/abs/1811.09600v3","url_pdf":"http://arxiv.org/pdf/1811.09600v3.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":"decoupling-direction-and-norm-for-efficient","repo_url":"https://github.com/jeromerony/adversarial-library","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"decoupling-direction-and-norm-for-efficient","repo_url":"https://github.com/jeromerony/fast_adversarial","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"BSD-3-Clause"}},{"paper_slug":"decoupling-direction-and-norm-for-efficient","repo_url":"https://github.com/anonICLR5/robust-interpretation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"decoupling-direction-and-norm-for-efficient","repo_url":"https://github.com/dwdavidxd/arn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"decoupling-direction-and-norm-for-efficient","repo_url":"https://github.com/luizgh/avc_nips_2018","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.09600","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1811.09600"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/anonICLR5/robust-interpretation","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/luizgh/avc_nips_2018","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jeromerony/fast_adversarial","reach":{"status":"ok","spdx":"BSD-3-Clause"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/jeromerony/adversarial-library","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dwdavidxd/arn","reach":{"status":"ok"}}],"summary":{"unverified":2},"by_repo_kind":{"named_in_paper":{"samples":2,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"57cb594ecfe4aa28","entry":"cosine_distance","repo":"jeromerony/fast_adversarial","repo_kind":"named_in_paper","path":"fast_adv/attacks/ddn_tf.py","file_url":"https://github.com/jeromerony/fast_adversarial/blob/HEAD/fast_adv/attacks/ddn_tf.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"57cb594ecfe4aa28"}},{"code_sha256_prefix":"3a2a9198bba99e2f","entry":"quantization","repo":"jeromerony/fast_adversarial","repo_kind":"named_in_paper","path":"fast_adv/attacks/ddn_tf.py","file_url":"https://github.com/jeromerony/fast_adversarial/blob/HEAD/fast_adv/attacks/ddn_tf.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"BSD-3-Clause","inline_ok":true,"mcp_get_code":{"code_sha256":"3a2a9198bba99e2f"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}