Intelligent Accompaniment Generation and Emotional Adaptation Using Graph Neural Networks to Model Chord Progression Logic

Intelligent Accompaniment Generation and Emotional Adaptation Using Graph Neural Networks to Model Chord Progression Logic | 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 Article Intelligent Accompaniment Generation and Emotional Adaptation Using Graph Neural Networks to Model Chord Progression Logic Suhao Bai This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8373845/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Current automatic accompaniment generation methods based on sequence models struggle to fully capture the long-term functional logic between chords, and their emotional adaptation accuracy is limited. To address this, this paper employs an intelligent accompaniment generation model based on a GNN (Graph Neural Network) and an EC-GNN-AG (Emotion-Conditioned Graph Neural Network for Accompaniment Generation) to explicitly model the functional logic between chords and achieve high-precision emotional control. This method constructs chord sequences as a multi-relational graph structure. The Graph Attention Network (GAT) is used to encode the harmonic functional relationships and surface affinities between nodes, and a cross-modal emotional attention mechanism is designed. Discrete or continuous emotional features are used as conditional inputs, achieving fine-grained emotional control through multi-level injection and dynamic attention weighting. Experiments show that EC-GNN-AG improves Harmonic Alignment Accuracy (HAA), which measures the alignment between the implied harmony of the generated accompaniment and the ground-truth chord sequence, by 12.6% compared to traditional LSTM-based models and improves emotional adaptability by 15.3% compared to the next-best Transformer-based model. In the Mean Opinion Score (MOS) test, the model significantly outperforms the comparison model in terms of harmonic pl ausibility (4.35 ± 0.12), emotional consistency (4.15 ± 0.14), and overall musicality (4.28 ± 0.11). EC-GNN-AG provides an effective solution for structure-aware and emotion-driven music generation. Physical sciences/Engineering Physical sciences/Mathematics and computing Graph Neural Network Intelligent Accompaniment Generation Emotional Adaptation Chord Progression Modeling Music Generation Multimodal Attention Mechanism Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 01 May, 2026 Reviewers agreed at journal 30 Apr, 2026 Reviews received at journal 30 Apr, 2026 Reviewers agreed at journal 28 Apr, 2026 Reviewers invited by journal 27 Apr, 2026 Editor assigned by journal 27 Apr, 2026 Editor invited by journal 19 Dec, 2025 Submission checks completed at journal 18 Dec, 2025 First submitted to journal 18 Dec, 2025 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-8373845","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":633616698,"identity":"a72b311c-93be-476c-a3e8-17ab700f7e79","order_by":0,"name":"Suhao Bai","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIie3PIWsDMRTA8RcepCZQe8foPsODwM0c61e5cRBdmJnrlYqZsrO5bxE4mE6JqCnUVp6qvqpODNZrayZGFlmRHwlE5M9LAKLoHiGuu4Lyx/FwZtVt+4kRL6mfKZlWwclYZKnu3YuxockUgR4EKdZu3OdRQz4xFg+ddwpCIQXlmG3Va2NASWP5E/2T2HKYwrO9kKyDywsFT/wJq5wgJ6S+Jj8hCbKFJpdQMiQGbEDiOEJPipLhL0xTKRvHM28yqnenr+I7n9fvrmWrt+fJx2Z58Ca/0XUBht6/JVEURdEfzrl5RbmX6aFMAAAAAElFTkSuQmCC","orcid":"","institution":"Chongqing College of Humanities，Science \u0026 Technology","correspondingAuthor":true,"prefix":"","firstName":"Suhao","middleName":"","lastName":"Bai","suffix":""}],"badges":[],"createdAt":"2025-12-16 08:54:41","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8373845/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8373845/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":108805723,"identity":"a38e6eae-2c89-49aa-bfe3-d5949e5179d1","added_by":"auto","created_at":"2026-05-08 15:26:43","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":658244,"visible":true,"origin":"","legend":"","description":"","filename":"revisedpaper.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8373845/v1_covered_98bafa34-6141-47bb-80a3-60462edc1b04.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Intelligent Accompaniment Generation and Emotional Adaptation Using Graph Neural Networks to Model Chord Progression Logic","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Graph Neural Network, Intelligent Accompaniment Generation, Emotional Adaptation, Chord Progression Modeling, Music Generation, Multimodal Attention Mechanism","lastPublishedDoi":"10.21203/rs.3.rs-8373845/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8373845/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCurrent automatic accompaniment generation methods based on sequence models struggle to fully capture the long-term functional logic between chords, and their emotional adaptation accuracy is limited. To address this, this paper employs an intelligent accompaniment generation model based on a GNN (Graph Neural Network) and an EC-GNN-AG (Emotion-Conditioned Graph Neural Network for Accompaniment Generation) to explicitly model the functional logic between chords and achieve high-precision emotional control. This method constructs chord sequences as a multi-relational graph structure. The Graph Attention Network (GAT) is used to encode the harmonic functional relationships and surface affinities between nodes, and a cross-modal emotional attention mechanism is designed. Discrete or continuous emotional features are used as conditional inputs, achieving fine-grained emotional control through multi-level injection and dynamic attention weighting. Experiments show that EC-GNN-AG improves Harmonic Alignment Accuracy (HAA), which measures the alignment between the implied harmony of the generated accompaniment and the ground-truth chord sequence, by 12.6% compared to traditional LSTM-based models and improves emotional adaptability by 15.3% compared to the next-best Transformer-based model. In the Mean Opinion Score (MOS) test, the model significantly outperforms the comparison model in terms of harmonic pl ausibility (4.35\u0026thinsp;\u0026plusmn;\u0026thinsp;0.12), emotional consistency (4.15\u0026thinsp;\u0026plusmn;\u0026thinsp;0.14), and overall musicality (4.28\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11). EC-GNN-AG provides an effective solution for structure-aware and emotion-driven music generation.\u003c/p\u003e","manuscriptTitle":"Intelligent Accompaniment Generation and Emotional Adaptation Using Graph Neural Networks to Model Chord Progression Logic","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-06 09:30:12","doi":"10.21203/rs.3.rs-8373845/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2026-05-01T09:04:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"140438140906287934597568997278227854939","date":"2026-04-30T10:35:48+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-30T08:20:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"329674277399098510761405920964827151394","date":"2026-04-28T06:33:00+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-28T02:35:58+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-04-27T11:22:28+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-12-19T11:43:41+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-12-18T09:36:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-12-18T09:20:04+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c0291338-f59d-42ed-9a5b-ffff72a7c3e6","owner":[],"postedDate":"May 6th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-01T09:04:42+00:00","index":84,"fulltext":""},{"type":"reviewerAgreed","content":"140438140906287934597568997278227854939","date":"2026-04-30T10:35:48+00:00","index":83,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-30T08:20:16+00:00","index":82,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":67437290,"name":"Physical sciences/Engineering"},{"id":67437291,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-05-06T09:30:12+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-06 09:30:12","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8373845","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8373845","identity":"rs-8373845","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}