ESRNN: Effective Residual Self-Attention Recurrent Neural Network with Soft Threading Function for Sound Event Location

ESRNN: Effective Residual Self-Attention Recurrent Neural Network with Soft Threading Function for Sound Event Location | 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 ESRNN: Effective Residual Self-Attention Recurrent Neural Network with Soft Threading Function for Sound Event Location Bin Zhang, Jiawen He, Peishun Liu, Liang Wang, Hao Zhou, Xuening Wang, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3767575/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 Sound event location is a critical aspect of two-dimensional direction-of-arrival (2D-DOA) estimation, predicting azimuth and elevation angles in 3D Cartesian coordinates for active sound events using multi-label regression. Challenges with conventional methods like the multi-signal classification (MUSIC) algorithm and baseline convolution recurrent neural network (BCRNN) include decreased precision and high computational demands, particularly in low signal-to-noise ratio (SNR) environments (SNR\textless-5 dB). Our work introduces an innovative solution, the effective residual self-attention recurrent neural network (ESRNN). ESRNN addresses distortion problems in low SNR conditions caused by the MUSIC algorithm, also enhancing 2D-DOA prediction accuracy in various SNR-reverberation scenarios. We propose two filter structures, ESRNN-L and ESRNN-G, tailored for SNRs above 0 dB and below -5 dB, respectively. Evaluating on TAU Spatial Sound Events 2019 datasets with synthetic SNRs from -10 dB to 30 dB, our experiments demonstrate ESRNN-L achieves a 21 \(%\) lower 2D-DOA error than BCRNN at SNRs below -5 dB. Additionally, ESRNN-G exhibits a 15 $ % $ lower error with a 10 $ % $ parameter reduction when SNRs exceed 0 dB. When compared with other principal attention methods through ablation study, it also showcases the model's efficiency and robustness. Sound event location Low signal-to-noise ratio 2D-DOA Residual self-attention Recurrent neural network Full Text 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-3767575","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":264746790,"identity":"9c61917d-1522-4eb4-ad4e-65cd8b282cc8","order_by":0,"name":"Bin Zhang","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Bin","middleName":"","lastName":"Zhang","suffix":""},{"id":264746791,"identity":"d5ba6a8d-ab12-4067-bb43-718e3206afde","order_by":1,"name":"Jiawen He","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Jiawen","middleName":"","lastName":"He","suffix":""},{"id":264746792,"identity":"657544f7-6058-4e56-90f4-0d571e6ad23c","order_by":2,"name":"Peishun Liu","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Peishun","middleName":"","lastName":"Liu","suffix":""},{"id":264746793,"identity":"5da082a6-1a47-4e35-bd62-e346af896fac","order_by":3,"name":"Liang Wang","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Liang","middleName":"","lastName":"Wang","suffix":""},{"id":264746794,"identity":"be818555-248c-47cd-aa41-46df06892f0b","order_by":4,"name":"Hao Zhou","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Hao","middleName":"","lastName":"Zhou","suffix":""},{"id":264746795,"identity":"a5adf65e-26f3-42cf-b5ac-4924f072ea88","order_by":5,"name":"Xuening Wang","email":"","orcid":"","institution":"Ocean University of China","correspondingAuthor":false,"prefix":"","firstName":"Xuening","middleName":"","lastName":"Wang","suffix":""},{"id":264746796,"identity":"45679351-9ae7-4a20-9c8a-eae272d66ffc","order_by":6,"name":"Ruichun Tang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4ElEQVRIiWNgGAWjYPACCQYDBgbGBxBOAn61PEhamA1I0cIA0sImQZQWe/azh19XtlnIm0vkHqv42GbHwM+eY8DwcwceW3jy0izPtkkY7pyRl3ZzZlsyg2TPGwPG3jP4HJZjZtjYJpFgcCPH7DZv2wEGIMOAmbENjxb+NwgtxSAt9gS1SOQYP4RpYQbbIkFIy403ZowN5yQMN5x5Yyw541wyj8SZZwUHe/FoYe/PMf7YUFYnb3A8x/DDhzI7Of725I0PfuLRwoCIDqi1IOIAXg3ASP9AQMEoGAWjYBSMdAAADbdKWbqsZFgAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-4273-8623","institution":"Ocean University of China","correspondingAuthor":true,"prefix":"","firstName":"Ruichun","middleName":"","lastName":"Tang","suffix":""}],"badges":[],"createdAt":"2023-12-17 14:24:30","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3767575/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3767575/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":68379255,"identity":"c90063da-c4ce-447b-b12e-9799227c34bf","added_by":"auto","created_at":"2024-11-06 16:07:19","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3494314,"visible":true,"origin":"","legend":"","description":"","filename":"JSIGD2300296.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3767575/v1_covered_034de913-0742-4a56-9533-4d4d24a8ad6e.pdf"}],"financialInterests":"","formattedTitle":"ESRNN: Effective Residual Self-Attention Recurrent Neural Network with Soft Threading Function for Sound Event Location","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":"[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":"Sound event location, Low signal-to-noise ratio, 2D-DOA, Residual self-attention, Recurrent neural network","lastPublishedDoi":"10.21203/rs.3.rs-3767575/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3767575/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eSound event location is a critical aspect of two-dimensional direction-of-arrival (2D-DOA) estimation, predicting azimuth and elevation angles in 3D Cartesian coordinates for active sound events using multi-label regression. Challenges with conventional methods like the multi-signal classification (MUSIC) algorithm and baseline convolution recurrent neural network (BCRNN) include decreased precision and high computational demands, particularly in low signal-to-noise ratio (SNR) environments (SNR\\textless-5 dB). Our work introduces an innovative solution, the effective residual self-attention recurrent neural network (ESRNN). ESRNN addresses distortion problems in low SNR conditions caused by the MUSIC algorithm, also enhancing 2D-DOA prediction accuracy in various SNR-reverberation scenarios. We propose two filter structures, ESRNN-L and ESRNN-G, tailored for SNRs above 0 dB and below -5 dB, respectively. Evaluating on TAU Spatial Sound Events 2019 datasets with synthetic SNRs from -10 dB to 30 dB, our experiments demonstrate ESRNN-L achieves a 21\u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e \\(%\\) \u003c/span\u003e \u003c/span\u003e lower 2D-DOA error than BCRNN at SNRs below -5 dB. Additionally, ESRNN-G exhibits a 15\u003cspan\u003e$\u003c/span\u003e%\u003cspan\u003e$\u003c/span\u003e lower error with a 10\u003cspan\u003e$\u003c/span\u003e%\u003cspan\u003e$\u003c/span\u003e parameter reduction when SNRs exceed 0 dB. When compared with other principal attention methods through ablation study, it also showcases the model's efficiency and robustness.\u003c/p\u003e","manuscriptTitle":"ESRNN: Effective Residual Self-Attention Recurrent Neural Network with Soft Threading Function for Sound Event Location","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-01-04 18:40:56","doi":"10.21203/rs.3.rs-3767575/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":"55ee894e-dee7-44dd-b65c-3debf0f92787","owner":[],"postedDate":"January 4th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-11-06T15:59:04+00:00","versionOfRecord":[],"versionCreatedAt":"2024-01-04 18:40:56","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3767575","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3767575","identity":"rs-3767575","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}