Dense-Stream YOLOv8n: A Lightweight Model Method for Real-Time Crowd Monitoring in Smart Libraries

Dense-Stream YOLOv8n: A Lightweight Model Method for Real-Time Crowd Monitoring in Smart Libraries | 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 Dense-Stream YOLOv8n: A Lightweight Model Method for Real-Time Crowd Monitoring in Smart Libraries Zini Chen, Xuming Xie, Taorong Qiu, Yuelei Yao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5446660/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Apr, 2025 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Crowd monitoring in the context of smart libraries is of great significance for resource optimization and service improvement. Addressing the challenges of insufficient accuracy and real-time performance in crowd detection under high-density and side-view scenarios in dynamic library environments, this paper proposes a crowd detection method based on an improved YOLOv8n model. First, side-view crowd videos from different time periods in the library were collected, segmented into images, and manually annotated to generate a high-quality training dataset. Then, a lightweight convolutional data augmentation module called DensityNet was designed to enhance the model's feature extraction capabilities in crowded occlusion scenes. Subsequently, model pruning and knowledge distillation techniques were combined to reduce model complexity and improve detection real-time performance, adapting it to the computational requirements of edge devices. Finally, a region detection algorithm was designed to better accommodate the needs of crowd monitoring in high-density and constrained-view dynamic environments by extending the detection trigger time, thereby providing an accurate and contactless solution for people flow monitoring in smart libraries. Experimental results show that the improved YOLOv8n model achieves an average precision ( [email protected] ) of 0.99 in high-density scenarios, close to the original model's 0.991, while achieving 0.861 in [email protected] :0.95, an increase of 0.014 over the model before pruning; in terms of real-time performance, the frame rate (FPS) significantly increased to 254, with computational costs reduced to 4.0 GFLOP and the number of parameters decreased to 2.04M, meeting the real-time detection needs of peak crowd environments in smart libraries. This study addresses the unique people flow monitoring requirements of smart libraries by proposing an efficient and accurate solution, which not only optimizes resource and service management but also provides new technical support for the intelligent management of other domains. Physical sciences/Mathematics and computing/Information technology Physical sciences/Mathematics and computing/Scientific data Physical sciences/Mathematics and computing/Computer science Smart Library YOLOv8n Model Pruning Knowledge Distillation Crowd Monitoring Full Text Additional Declarations No competing interests reported. Supplementary Files Supplementary.docx video.rar Cite Share Download PDF Status: Published Journal Publication published 04 Apr, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 26 Nov, 2024 Reviews received at journal 24 Nov, 2024 Reviewers agreed at journal 24 Nov, 2024 Reviews received at journal 24 Nov, 2024 Reviewers agreed at journal 23 Nov, 2024 Reviewers invited by journal 23 Nov, 2024 Editor assigned by journal 21 Nov, 2024 Editor invited by journal 15 Nov, 2024 Submission checks completed at journal 14 Nov, 2024 First submitted to journal 13 Nov, 2024 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-5446660","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":382855064,"identity":"e9280d0a-0772-412f-8b7c-6b04b6677a86","order_by":0,"name":"Zini Chen","email":"","orcid":"","institution":"Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Zini","middleName":"","lastName":"Chen","suffix":""},{"id":382855073,"identity":"a62972bb-06b4-4581-81e7-c9eee31c3368","order_by":1,"name":"Xuming Xie","email":"","orcid":"","institution":"Library of Nanchang University","correspondingAuthor":false,"prefix":"","firstName":"Xuming","middleName":"","lastName":"Xie","suffix":""},{"id":382855074,"identity":"1ac915c1-04d8-43ad-b60a-f9c56425e7e2","order_by":2,"name":"Taorong Qiu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA3UlEQVRIiWNgGAWjYDACZgTrAIQ+QFgLYwOExZZApBYGuBYeA+K08B3nMX/wo+KenDn/mm8PfrYxyPHdSGD8XIBHi+RhHsPGnjPFxpYz3m437G1jMJa8kcAsPQOPFgOglgbetoTEDTfObpNmbGMAMhLYmHkIaGn8C9Zy5hlISz1RWprBtpzvYQNpSTAgpEXyMFvhbJkzCcYGN9jMJHvOSRjOPPOwWRqfFr7zhzd8fFORIGdw/vAziR9lNvJ8x5MPfsanBRELEglgkgERUQS18B/ArWgUjIJRMApGNgAAGqZN/P8zKGEAAAAASUVORK5CYII=","orcid":"","institution":"Nanchang University","correspondingAuthor":true,"prefix":"","firstName":"Taorong","middleName":"","lastName":"Qiu","suffix":""},{"id":382855075,"identity":"5056dde2-c945-4754-b92e-c81345369191","order_by":3,"name":"Yuelei Yao","email":"","orcid":"","institution":"Jiangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yuelei","middleName":"","lastName":"Yao","suffix":""}],"badges":[],"createdAt":"2024-11-13 11:38:33","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5446660/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5446660/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-94659-x","type":"published","date":"2025-04-04T15:57:39+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80082252,"identity":"a581533c-3a0b-4934-87ac-05d5aab794ca","added_by":"auto","created_at":"2025-04-07 16:08:02","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":933921,"visible":true,"origin":"","legend":"","description":"","filename":"Manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5446660/v1_covered_6457a8ed-f0ea-403b-a157-93d8ae4a0013.pdf"},{"id":71198278,"identity":"1a3292b5-76fb-4858-968f-746788af9776","added_by":"auto","created_at":"2024-12-12 05:41:08","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":49868834,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-5446660/v1/e550a9bfacbae3d60b112e1b.docx"},{"id":71198610,"identity":"ab43f917-d06d-4cb0-9353-e4e961648187","added_by":"auto","created_at":"2024-12-12 05:57:40","extension":"rar","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":49712724,"visible":true,"origin":"","legend":"","description":"","filename":"video.rar","url":"https://assets-eu.researchsquare.com/files/rs-5446660/v1/bbca2e75eb5ba04acc2325d2.rar"}],"financialInterests":"No competing interests reported.","formattedTitle":"Dense-Stream YOLOv8n: A Lightweight Model Method for Real-Time Crowd Monitoring in Smart Libraries","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"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":"Smart Library, YOLOv8n, Model Pruning, Knowledge Distillation, Crowd Monitoring","lastPublishedDoi":"10.21203/rs.3.rs-5446660/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5446660/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCrowd monitoring in the context of smart libraries is of great significance for resource optimization and service improvement. Addressing the challenges of insufficient accuracy and real-time performance in crowd detection under high-density and side-view scenarios in dynamic library environments, this paper proposes a crowd detection method based on an improved YOLOv8n model. First, side-view crowd videos from different time periods in the library were collected, segmented into images, and manually annotated to generate a high-quality training dataset. Then, a lightweight convolutional data augmentation module called DensityNet was designed to enhance the model's feature extraction capabilities in crowded occlusion scenes. Subsequently, model pruning and knowledge distillation techniques were combined to reduce model complexity and improve detection real-time performance, adapting it to the computational requirements of edge devices. Finally, a region detection algorithm was designed to better accommodate the needs of crowd monitoring in high-density and constrained-view dynamic environments by extending the detection trigger time, thereby providing an accurate and contactless solution for people flow monitoring in smart libraries. Experimental results show that the improved YOLOv8n model achieves an average precision ([email protected]) of 0.99 in high-density scenarios, close to the original model's 0.991, while achieving 0.861 in [email protected]:0.95, an increase of 0.014 over the model before pruning; in terms of real-time performance, the frame rate (FPS) significantly increased to 254, with computational costs reduced to 4.0 GFLOP and the number of parameters decreased to 2.04M, meeting the real-time detection needs of peak crowd environments in smart libraries. This study addresses the unique people flow monitoring requirements of smart libraries by proposing an efficient and accurate solution, which not only optimizes resource and service management but also provides new technical support for the intelligent management of other domains.\u003c/p\u003e","manuscriptTitle":"Dense-Stream YOLOv8n: A Lightweight Model Method for Real-Time Crowd Monitoring in Smart Libraries","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-12 05:40:59","doi":"10.21203/rs.3.rs-5446660/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-11-26T11:51:40+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-24T09:41:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"208451788180067610626442517697613116750","date":"2024-11-24T09:02:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-24T08:08:08+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"2395916079456536031192377843775620983","date":"2024-11-23T16:58:36+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-23T13:44:57+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-22T04:22:21+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-11-15T06:58:22+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-14T09:52:08+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-11-13T11:34:01+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":"59f7337b-e9f5-4ced-8fe2-4c15b47dc1ad","owner":[],"postedDate":"December 12th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":40776612,"name":"Physical sciences/Mathematics and computing/Information technology"},{"id":40776614,"name":"Physical sciences/Mathematics and computing/Scientific data"},{"id":40776615,"name":"Physical sciences/Mathematics and computing/Computer science"}],"tags":[],"updatedAt":"2025-04-07T16:04:30+00:00","versionOfRecord":{"articleIdentity":"rs-5446660","link":"https://doi.org/10.1038/s41598-025-94659-x","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-04-04 15:57:39","publishedOnDateReadable":"April 4th, 2025"},"versionCreatedAt":"2024-12-12 05:40:59","video":"","vorDoi":"10.1038/s41598-025-94659-x","vorDoiUrl":"https://doi.org/10.1038/s41598-025-94659-x","workflowStages":[]},"version":"v1","identity":"rs-5446660","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5446660","identity":"rs-5446660","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}