C2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification

C2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification | 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 C2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification Nur Alam, Md Rakibul Islam, L. Minh Dang, Mariya Kibtiya Kibtiya, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6339269/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 Few-shot fine-grained image classification (FS-FGIC) intends on improving the ability to classify detailed image categories with limited training samples. However, two major challenges still exist. The challenges include effectively extracting the essential features needed for fine-grained classification while minimizing irrelevant noise, which can cause overfitting when dealing with few-shot conditions. The second challenge lies in achieving robust feature alignment between the support and query samples, especially when there are spatial variations, such as differences in the positions or angles of the objects. This paper introduces C2-Net to address these issues. This innovative framework includes two key modules designed to overcome these challenges. The Cross-Layer Feature Refinement (CLFR) module has an impact on the quality of features. It does this by blending outputs from several layers of the network. This approach helps to cut down noise at the sample level. At the same time, the Cross-Sample Feature Adjustment (CSFA) module changes to fit spatial and channel differences. This makes sure that features line up between the few support and query samples. Through these mechanisms, C2-Net reduces misalignments and improves feature discrimination. Comprehensive experiments conducted on five benchmark datasets demonstrate that C2-Net continously exceeds existing methods, achieving state-of-the-art (SOTA) results in most cases, such as improved One-shot classification accuracy on the CUB dataset from 54.87% to 76.51% and 5-shot accuracy from 79.09% to 88.15%. This approach represents a significant advancement in tackling the challenges of FS-FGIC. Quadrupole exciton Polariton WGM BEC 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-6339269","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":437182846,"identity":"ecc6c99d-a6c0-43a4-a1ac-55ad2e044a25","order_by":0,"name":"Nur Alam","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYJCCw3//2MgxMBwAcwyI0cH4gLchzZgkLcwGvA2HEhugPMJazCVyn0lI7jiQPr/xjAHDjxoGY/MGAlosZ6SbSRieuZO74cAZA8aeYwxmMgcIaDG4kcYmkcD2LHcDA9AW3gYGGwlCDgNrOcB2OF2+AWjLXyK1MBs2th1OYAA6jBloixlhLWeeMT5mOJNmuOHAsYLDMsckjAlrOZ7GcJihwkZefsbhjQ/f1NgYziCkBQEkDoAik6AdyIC/gRTVo2AUjIJRMJIAAFBpQP02UpqkAAAAAElFTkSuQmCC","orcid":"","institution":"Sejong University","correspondingAuthor":true,"prefix":"","firstName":"Nur","middleName":"","lastName":"Alam","suffix":""},{"id":437182848,"identity":"ef948252-3e12-40fe-8030-80e6a99b02da","order_by":1,"name":"Md Rakibul Islam","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Md","middleName":"Rakibul","lastName":"Islam","suffix":""},{"id":437182850,"identity":"1f8fd975-7e33-44d8-9644-ed6dc36b3d62","order_by":2,"name":"L. Minh Dang","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"L.","middleName":"Minh","lastName":"Dang","suffix":""},{"id":437182851,"identity":"778ec36c-11aa-4306-967c-1adeb477f970","order_by":3,"name":"Mariya Kibtiya Kibtiya","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Mariya","middleName":"Kibtiya","lastName":"Kibtiya","suffix":""},{"id":437182852,"identity":"49c71382-d566-4513-a83f-35654c2d5f1a","order_by":4,"name":"Adnan Hossan Rifat","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Adnan","middleName":"Hossan","lastName":"Rifat","suffix":""},{"id":437182853,"identity":"f87870c0-77a1-4766-bf0a-f9b5fe96e8af","order_by":5,"name":"Moon Hyeonjoon","email":"","orcid":"","institution":"Sejong University","correspondingAuthor":false,"prefix":"","firstName":"Moon","middleName":"","lastName":"Hyeonjoon","suffix":""}],"badges":[],"createdAt":"2025-03-30 15:23:19","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6339269/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6339269/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":81654650,"identity":"10571edc-0c27-4890-adbc-24d5321b27fb","added_by":"auto","created_at":"2025-04-29 17:46:27","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":667494,"visible":true,"origin":"","legend":"","description":"","filename":"C2Net.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6339269/v1_covered_ee890011-cd19-4722-ad83-eb29e81210f1.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eC2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"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":"[email protected]","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":"Quadrupole exciton, Polariton, WGM, BEC","lastPublishedDoi":"10.21203/rs.3.rs-6339269/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6339269/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eFew-shot fine-grained image classification (FS-FGIC) intends on improving the ability to classify detailed image categories with limited training samples. However, two major challenges still exist. The challenges include effectively extracting the essential features needed for fine-grained classification while minimizing irrelevant noise, which can cause overfitting when dealing with few-shot conditions. The second challenge lies in achieving robust feature alignment between the support and query samples, especially when there are spatial variations, such as differences in the positions or angles of the objects. This paper introduces C2-Net to address these issues. This innovative framework includes two key modules designed to overcome these challenges. The Cross-Layer Feature Refinement (CLFR) module has an impact on the quality of features. It does this by blending outputs from several layers of the network. This approach helps to cut down noise at the sample level. At the same time, the Cross-Sample Feature Adjustment (CSFA) module changes to fit spatial and channel differences. This makes sure that features line up between the few support and query samples. Through these mechanisms, C2-Net reduces misalignments and improves feature discrimination. Comprehensive experiments conducted on five benchmark datasets demonstrate that C2-Net continously exceeds existing methods, achieving state-of-the-art (SOTA) results in most cases, such as improved One-shot classification accuracy on the CUB dataset from 54.87% to 76.51% and 5-shot accuracy from 79.09% to 88.15%. This approach represents a significant advancement in tackling the challenges of FS-FGIC.\u003c/p\u003e","manuscriptTitle":"C2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-23 06:04:06","doi":"10.21203/rs.3.rs-6339269/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","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":"3bd7f96f-aad6-403a-9550-42a6e527fdac","owner":[],"postedDate":"April 23rd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-04-29T17:38:20+00:00","versionOfRecord":[],"versionCreatedAt":"2025-04-23 06:04:06","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6339269","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6339269","identity":"rs-6339269","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}