{"paper_id":"11185635-b719-476e-8cd7-e46f629cea36","body_text":"SP-LID:Subtle Perturbation Sensitive Adversarial Example Detection Method Based on Local Intrinsic Dimension | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article SP-LID:Subtle Perturbation Sensitive Adversarial Example Detection Method Based on Local Intrinsic Dimension JiaWei Ge, Juan Wang, Yue Yu, Ran Chen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4978361/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Computer vision models based on deep learning technology are vulnerable to adversarial examples. By adding some subtle perturbations to the examples, the attacker can make the deep learning model make mistakes, which will lead to serious consequences. In order to better defend against this attack, one of the methods is to detect and cull the adversarial examples. Compared with the original local intrinsic dimension detection method, this paper proposes an optimized local intrinsic dimension detection method to characterize the dimensional properties of adversarial examples. This method not only detects the distance distribution of a example to its neighbors, but also evaluates the sensitivity of a example to perturbations to determine whether it is an adversarial example. Four different adversarial attack strategies were used to evaluate the defense effect of the proposed method. The experimental results show that the improved local intrinsic dimension detection method is more effective than other defense methods, and plays a significant role in different data sets. Deep Learning Adversarial Attack Adversarial Examples Local Intrinsic Dimensionality Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-4978361\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":358914454,\"identity\":\"9569fcb9-e893-4c40-96cd-4ff3b83a504c\",\"order_by\":0,\"name\":\"JiaWei Ge\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Chengdu University of Information Technology\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"JiaWei\",\"middleName\":\"\",\"lastName\":\"Ge\",\"suffix\":\"\"},{\"id\":358914455,\"identity\":\"f6f8004e-02b2-48b5-b60b-4d1eb5102d8f\",\"order_by\":1,\"name\":\"Juan Wang\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA30lEQVRIiWNgGAWjYLACCSDmB+IDQMzYQLQWSaDSAweI1gICBgcg1hDWYnD87OEXlm12ecbHzxge/sBgI7vhAPOzB3i1nMlLs5BsSy42O5NjAHRYmvGGA2zmBvi0mB3IMTOQbGNO3HaDB6TlcOKGAzxsEni1nH8D0lKfuHkGWMt/IrTcyDF+INkGNFwCrOUAYS32N96YMUicO54440xawYEzBsnGMw+zmeHVItmfY/xZoqw6sb/98OYPFRV2sn3Hm5/h1QIEbNIIFaCgYiagHqTk4wfCikbBKBgFo2AkAwBZt0/AFV4ipgAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"Chengdu University of Information Technology\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Juan\",\"middleName\":\"\",\"lastName\":\"Wang\",\"suffix\":\"\"},{\"id\":358914456,\"identity\":\"a2dda4de-0082-445c-85e8-dd07a638e8a8\",\"order_by\":2,\"name\":\"Yue Yu\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Chengdu University of Information Technology\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yue\",\"middleName\":\"\",\"lastName\":\"Yu\",\"suffix\":\"\"},{\"id\":358914458,\"identity\":\"267ec4c7-71ca-43e2-86fb-cfac41f0344d\",\"order_by\":3,\"name\":\"Ran Chen\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Chengdu University of Information Technology\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Ran\",\"middleName\":\"\",\"lastName\":\"Chen\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2024-08-26 13:38:55\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-4978361/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-4978361/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":67604675,\"identity\":\"34074eb2-8c3e-4c99-9e47-786893fc74e4\",\"added_by\":\"auto\",\"created_at\":\"2024-10-28 04:17:18\",\"extension\":\"pdf\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":875026,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"SPLIDSubtlePerturbationSensitiveAdversarialExampleDetectionMethodBasedonLocalIntrinsicDimension.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-4978361/v1_covered_cdae3f89-839f-46ee-a5aa-fb23ebdb3736.pdf\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"SP-LID:Subtle Perturbation Sensitive Adversarial Example Detection Method Based on Local Intrinsic Dimension\",\"fulltext\":[],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":false,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":true,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":true,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Deep Learning, Adversarial Attack, Adversarial Examples, Local Intrinsic Dimensionality\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-4978361/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-4978361/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003eComputer vision models based on deep learning technology are vulnerable to adversarial examples. By adding some subtle perturbations to the examples, the attacker can make the deep learning model make mistakes, which will lead to serious consequences. In order to better defend against this attack, one of the methods is to detect and cull the adversarial examples. Compared with the original local intrinsic dimension detection method, this paper proposes an optimized local intrinsic dimension detection method to characterize the dimensional properties of adversarial examples. This method not only detects the distance distribution of a example to its neighbors, but also evaluates the sensitivity of a example to perturbations to determine whether it is an adversarial example. Four different adversarial attack strategies were used to evaluate the defense effect of the proposed method. The experimental results show that the improved local intrinsic dimension detection method is more effective than other defense methods, and plays a significant role in different data sets.\\u003c/p\\u003e\",\"manuscriptTitle\":\"SP-LID:Subtle Perturbation Sensitive Adversarial Example Detection Method Based on Local Intrinsic Dimension\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2024-09-26 04:12:33\",\"doi\":\"10.21203/rs.3.rs-4978361/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"478f8237-9057-415b-b586-b2ba4064f746\",\"owner\":[],\"postedDate\":\"September 26th, 2024\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2024-10-28T04:08:59+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2024-09-26 04:12:33\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-4978361\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-4978361\",\"identity\":\"rs-4978361\",\"version\":[\"v1\"]},\"buildId\":\"qtupq5eGEP_6zYnWcrvyt\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}