SAREC: A Semantic- Aware Retrieval-Augmented Conformer for Multilingual Low-Resource Speech Recognition

SAREC: A Semantic- Aware Retrieval-Augmented Conformer for Multilingual Low-Resource Speech Recognition | 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 SAREC: A Semantic- Aware Retrieval-Augmented Conformer for Multilingual Low-Resource Speech Recognition B. Arukiran Reddy, S. Udaya Bhaskar, J. Sunil Kumar, P. Raghunadh This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8413645/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 Low-resource speech recognition remains fundamentally limited by training distribution, as neural networks cannot access linguistic knowledge beyond their training data. SAREC (Semantic-Aware Retrieval Enhanced Conformer), a novel neural architecture that augments Conformer-based acoustic encoders with dynamically retrieved phonetic exemplars during both training and inference. The key innovation is frame-level retrieval with learned cross-attention fusion, enabling phonetic-precision knowledge integration at each encoding layer. Unlike existing utterance-level retrieval approaches (SRAG), SAREC system retrieves and fuses acoustic prototypes at frame granularity, capturing phonetic patterns at multiple scales through strategic insertion at encoder layers 3, 6, 9, and 12. A learned gating mechanism dynamically weights acoustic versus exemplar information per frame, providing principled information fusion superior to simple concatenation. Evaluation on Telugu, Kannada, Tamil demonstrates consistent improvements: Telugu 18.2% WER (Word Error Rate) ( (vs. 21.4% baseline, 15.0% relative, p < 0.001, 95% CI: [17.8, 18.6]%); Kannada 20.3% (vs. 24.8%, 18.1% relative); Tamil 16.4% (vs. 19.2%, 14.6% relative). Robustness improves 12.1% at 5 dB SNR (Signal to Noise Ratio), latency remains practical (356 ms/10s audio). Ablation studies confirm frame-level superiority (15% over utterance-level, K = 5 optimal).Three mechanisms explain gains: (1) phonological disambiguation, (2) morphological integration, (3) cross-dialect generalization. Results advance low-resource ASR accessibility through systematic frame-level retrieval augmentation Low-resource speech recognition Retrieval augmentation Exemplar-based learning Knowledge-grounded neural networks Acoustic modeling 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-8413645","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":565405028,"identity":"33698a65-1cd7-40f1-ba15-f6966c31e118","order_by":0,"name":"B. Arukiran Reddy","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+ElEQVRIiWNgGAWjYHACZoYECMNAgqECxGduIEXLGRCfkQgtDDAtjG0gmoAW/tlnjA0e7rHLM29v3nibd15tNH87UMuPim04tUicyzFOSHiWXCxz5lixNe+247kzDjM2MPacuY3bmjM8xgcSDjAnzpDIMZPm3XYstwGohZmxDbcWeYiW+sQZ8m+AWuYcy51PSIsBUEtCwoHDQFt4gFoaanI3ENJieIat2CDhwPHEGTxpxZZzjh3I3QjUchCfX+TOMG+W/HGgOnEG++GNN97U1OXOO3/44IMfFXi8jwyYeBgOgxkHiFMPBIw/GOqIVjwKRsEoGAUjBwAA9kdbbgRGCv0AAAAASUVORK5CYII=","orcid":"","institution":"Jawaharlal Nehru Technological University, Hyderabad","correspondingAuthor":true,"prefix":"","firstName":"B.","middleName":"Arukiran","lastName":"Reddy","suffix":""},{"id":565405029,"identity":"9e793e24-0296-4cea-9478-e7aea397073c","order_by":1,"name":"S. Udaya Bhaskar","email":"","orcid":"","institution":"Jawaharlal Nehru Technological University, Hyderabad","correspondingAuthor":false,"prefix":"","firstName":"S.","middleName":"Udaya","lastName":"Bhaskar","suffix":""},{"id":565405030,"identity":"c79fbca8-c395-4f1e-9718-9693ed014ef9","order_by":2,"name":"J. Sunil Kumar","email":"","orcid":"","institution":"Jawaharlal Nehru Technological University, Hyderabad","correspondingAuthor":false,"prefix":"","firstName":"J.","middleName":"Sunil","lastName":"Kumar","suffix":""},{"id":565405031,"identity":"0278dcb3-9f35-4443-96b6-19418a33e024","order_by":3,"name":"P. Raghunadh","email":"","orcid":"","institution":"Jawaharlal Nehru Technological University, Hyderabad","correspondingAuthor":false,"prefix":"","firstName":"P.","middleName":"","lastName":"Raghunadh","suffix":""}],"badges":[],"createdAt":"2025-12-20 17:38:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8413645/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8413645/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99315980,"identity":"06bd00cd-187d-41ea-a8cd-61dc9170cad1","added_by":"auto","created_at":"2025-12-31 16:27:32","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":756700,"visible":true,"origin":"","legend":"","description":"","filename":"FinalDraftSARECPaperversion1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8413645/v1_covered_449c63cf-ed97-44e2-b8d2-0dd543ca7080.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"SAREC: A Semantic- Aware Retrieval-Augmented Conformer for Multilingual Low-Resource Speech Recognition","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":"Low-resource speech recognition, Retrieval augmentation, Exemplar-based learning, Knowledge-grounded neural networks, Acoustic modeling","lastPublishedDoi":"10.21203/rs.3.rs-8413645/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8413645/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eLow-resource speech recognition remains fundamentally limited by training distribution, as neural networks cannot access linguistic knowledge beyond their training data. SAREC (Semantic-Aware Retrieval Enhanced Conformer), a novel neural architecture that augments Conformer-based acoustic encoders with dynamically retrieved phonetic exemplars during both training and inference. The key innovation is frame-level retrieval with learned cross-attention fusion, enabling phonetic-precision knowledge integration at each encoding layer. Unlike existing utterance-level retrieval approaches (SRAG), SAREC system retrieves and fuses acoustic prototypes at frame granularity, capturing phonetic patterns at multiple scales through strategic insertion at encoder layers 3, 6, 9, and 12. A learned gating mechanism dynamically weights acoustic versus exemplar information per frame, providing principled information fusion superior to simple concatenation. Evaluation on Telugu, Kannada, Tamil demonstrates consistent improvements: Telugu 18.2% WER (Word Error Rate) ( (vs. 21.4% baseline, 15.0% relative, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, 95% CI: [17.8, 18.6]%); Kannada 20.3% (vs. 24.8%, 18.1% relative); Tamil 16.4% (vs. 19.2%, 14.6% relative). Robustness improves 12.1% at 5 dB SNR (Signal to Noise Ratio), latency remains practical (356 ms/10s audio). Ablation studies confirm frame-level superiority (15% over utterance-level, K\u0026thinsp;=\u0026thinsp;5 optimal).Three mechanisms explain gains: (1) phonological disambiguation, (2) morphological integration, (3) cross-dialect generalization. Results advance low-resource ASR accessibility through systematic frame-level retrieval augmentation\u003c/p\u003e","manuscriptTitle":"SAREC: A Semantic- Aware Retrieval-Augmented Conformer for Multilingual Low-Resource Speech Recognition","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-29 15:00:53","doi":"10.21203/rs.3.rs-8413645/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":"186211f4-4ecd-471e-8f0b-f0a3c801fca4","owner":[],"postedDate":"December 29th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-12-29T15:00:53+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-29 15:00:53","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8413645","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8413645","identity":"rs-8413645","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}