DMLNet:Densely Connected and Multi-Scale Lightweight High-Resolution Human Pose Estimation Network | 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 DMLNet:Densely Connected and Multi-Scale Lightweight High-Resolution Human Pose Estimation Network Chunsheng Zhang, Wanggen Li, Cheng Wang, Yuchen Li, Shangshu Gao This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7739102/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 The current research on human pose estimation mainly focuses on improving model accuracy, while efficiency optimization in resource limited scenarios such as mobile devices or edge devices has not been fully explored. Specifically, how to maintain high accuracy under low parameter count ( 30FPS) conditions remains a problem.This paper introduces a densely connected and multi-scale lightweight high-resolution human pose estimation network, termed DMLNet, which demonstrates superior performance in comparison to existing lightweight networks.DMLNet introduces the Stem module, TMCneck module, and KTConv module based on HRNet(high-resolution network). The Stem module overcomes the limitations of the original HRNet modules, which could only extract basic information from feature maps, enabling the capture of partial details and deeper information from these maps. The TMCneck module, with its innovative network structure and incorporation of attention mechanisms, significantly enhances the model's accuracy. Meanwhile, the KTConv module achieves both model lightweighting and the extraction of feature information across multiple scales.Additionally, we have introduced dense connections and integrated a feature fusion strategy that combines RFCA attention to enhance the model's performance.We conducted extensive experiments on the COCO and MPII validation datasets, and the results significantly surpass those of existing networks. Lightweight model Human pose estimation High-resolution network Dense connections Multi-scale feature fusion 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-7739102","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":533051098,"identity":"00d3c30c-68c9-495e-9316-898f6652a566","order_by":0,"name":"Chunsheng Zhang","email":"","orcid":"","institution":"Anhui Normal University","correspondingAuthor":false,"prefix":"","firstName":"Chunsheng","middleName":"","lastName":"Zhang","suffix":""},{"id":533051099,"identity":"75ba4b60-9124-470c-b653-c94c279c6163","order_by":1,"name":"Wanggen Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAvElEQVRIiWNgGAWjYFACHiCuYEiAs4nUcoZkLYxtpGgxOH724OPCeXZ55hIJjA/etjHImxPSItmTl2w8c1tyseWMBGbDuW0MhjsbCGjhl+Axk+bdxpy44UYCmzQv0IUGBwhoYZPgMf/NO6cepIX9N1FaQLYw8zYcBtvCTJQWyZ4cY2meY8cTN5x52Cw555yE4QZCWgyOnzH8zFNTnbjhePLBD2/KbOQJ2oIEGBuAhATx6kfBKBgFo2AU4AYA7087OxbNWCsAAAAASUVORK5CYII=","orcid":"","institution":"Anhui Normal University","correspondingAuthor":true,"prefix":"","firstName":"Wanggen","middleName":"","lastName":"Li","suffix":""},{"id":533051100,"identity":"d580a2bf-379c-46e5-ab10-9872d076c798","order_by":2,"name":"Cheng Wang","email":"","orcid":"","institution":"Anhui Normal University","correspondingAuthor":false,"prefix":"","firstName":"Cheng","middleName":"","lastName":"Wang","suffix":""},{"id":533051101,"identity":"84d3b39e-6a4d-4698-b736-3a2c7b7d45e6","order_by":3,"name":"Yuchen Li","email":"","orcid":"","institution":"Anhui Normal University","correspondingAuthor":false,"prefix":"","firstName":"Yuchen","middleName":"","lastName":"Li","suffix":""},{"id":533051102,"identity":"e603d76d-d330-4291-87d9-2b642436a9cf","order_by":4,"name":"Shangshu Gao","email":"","orcid":"","institution":"Anhui Normal University","correspondingAuthor":false,"prefix":"","firstName":"Shangshu","middleName":"","lastName":"Gao","suffix":""}],"badges":[],"createdAt":"2025-09-29 07:23:31","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7739102/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7739102/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94382509,"identity":"ed568a10-61f4-4f85-8604-63512a943397","added_by":"auto","created_at":"2025-10-27 13:45:15","extension":"doc","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4670538,"visible":true,"origin":"","legend":"","description":"","filename":"DML.doc","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/ddda71504ab2c0b1362dba1b.doc"},{"id":94383136,"identity":"796fd811-4d5b-4925-ad5e-d1a8473fecec","added_by":"auto","created_at":"2025-10-27 13:45:33","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6406,"visible":true,"origin":"","legend":"","description":"","filename":"193cc26c04ca45188bbd1196034947d9.json","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/453ccaa240b7d7eeab5a6292.json"},{"id":94383171,"identity":"00839166-514f-453e-a3e2-61afc23d3eba","added_by":"auto","created_at":"2025-10-27 13:45:39","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121661,"visible":true,"origin":"","legend":"","description":"","filename":"193cc26c04ca45188bbd1196034947d91enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/360a8e995cda56214ff8f58e.xml"},{"id":94383681,"identity":"5af77d15-21e6-49cd-a41d-7f65d4ce25fd","added_by":"auto","created_at":"2025-10-27 13:46:44","extension":"emf","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2817328,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.emf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/d71fb57508a6d4bf55807481.emf"},{"id":94383399,"identity":"8c19fe23-40db-4d86-8f74-aa1f807e3936","added_by":"auto","created_at":"2025-10-27 13:46:05","extension":"emf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":231172,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.emf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/c790fbeefbefc695c1d78def.emf"},{"id":94383407,"identity":"4ccda120-379e-4d43-837c-4ea2688189f9","added_by":"auto","created_at":"2025-10-27 13:46:06","extension":"emf","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":291360,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/74dc059d1c14f5bc6c4ce78b.emf"},{"id":94383611,"identity":"f23646d9-b247-4334-a81e-46393949c225","added_by":"auto","created_at":"2025-10-27 13:46:37","extension":"emf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":808904,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.emf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/8ec7254c5a091a568c319799.emf"},{"id":94383264,"identity":"05710a69-43f2-4426-b40a-5a7fdfba539f","added_by":"auto","created_at":"2025-10-27 13:45:46","extension":"emf","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53484,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/9beba75a84ad9f710b383a1a.emf"},{"id":94383683,"identity":"0eb509b3-0683-4a62-9aab-89a29bc30db1","added_by":"auto","created_at":"2025-10-27 13:46:44","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":490312,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/684f3e978e99b9b1afa1b321.png"},{"id":94383432,"identity":"5c2dff9c-35a0-4313-94de-c2b26e4bab0d","added_by":"auto","created_at":"2025-10-27 13:46:11","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":389893,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/a00afdfc7cd590de4d361f56.jpeg"},{"id":94383168,"identity":"fe7c6edc-96db-4e06-af6b-5ecf627fe12d","added_by":"auto","created_at":"2025-10-27 13:45:38","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":119343,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/1873f95ab0121713fbbf970f.png"},{"id":94383427,"identity":"e59ca11f-6eec-4997-aa15-bb4523602f28","added_by":"auto","created_at":"2025-10-27 13:46:10","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7804,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/b7813769bfeaae61816c6493.png"},{"id":94381849,"identity":"e7494785-d4f7-4659-9099-9b8cd35b3f6f","added_by":"auto","created_at":"2025-10-27 13:44:20","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":13122,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/ec2e86b159888ed07f731ac8.png"},{"id":94383433,"identity":"78d5fec3-658d-49da-92c1-1c41589670ac","added_by":"auto","created_at":"2025-10-27 13:46:12","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12601,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/114405a3240c5c8f90ff1695.png"},{"id":94383383,"identity":"1e77d311-7fc5-4cb8-9648-0047a3951cb1","added_by":"auto","created_at":"2025-10-27 13:45:59","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":7397,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/cd526fe7caabd54dc25ed0d1.png"},{"id":94382670,"identity":"1c80c640-9cb6-4e36-ac29-ec14d19beece","added_by":"auto","created_at":"2025-10-27 13:45:29","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":87556,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/52dafa3633d80f067e4cd8bf.png"},{"id":94383169,"identity":"dc62e95c-5b28-4ea5-9426-d897ee72244f","added_by":"auto","created_at":"2025-10-27 13:45:39","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":115316,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/d955402568fb66c60bf96b94.png"},{"id":94383449,"identity":"a0119209-3232-4673-984b-fe64fbb787cd","added_by":"auto","created_at":"2025-10-27 13:46:14","extension":"xml","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":121818,"visible":true,"origin":"","legend":"","description":"","filename":"193cc26c04ca45188bbd1196034947d91structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/0dcdf62c7feeed12a381c80b.xml"},{"id":94383139,"identity":"f50d586e-094c-4ade-99b4-7d09cdc16c3c","added_by":"auto","created_at":"2025-10-27 13:45:33","extension":"html","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":133304,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1/b85af5c913946cd03df5598f.html"},{"id":104403855,"identity":"fe780b99-ea8e-4224-aecb-46f4c3a5bb69","added_by":"auto","created_at":"2026-03-11 12:19:13","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":690467,"visible":true,"origin":"","legend":"","description":"","filename":"DML.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7739102/v1_covered_d0c1a12d-b503-4e2c-9b6e-1b1a71e10237.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eDMLNet:Densely Connected and Multi-Scale Lightweight High-Resolution Human Pose Estimation Network\u003c/p\u003e","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":"Lightweight model, Human pose estimation, High-resolution network, Dense connections, Multi-scale feature fusion","lastPublishedDoi":"10.21203/rs.3.rs-7739102/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7739102/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe current research on human pose estimation mainly focuses on improving model accuracy, while efficiency optimization in resource limited scenarios such as mobile devices or edge devices has not been fully explored. Specifically, how to maintain high accuracy under low parameter count (\u0026lt;\u0026thinsp;5M) and real-time inference (\u0026gt;\u0026thinsp;30FPS) conditions remains a problem.This paper introduces a densely connected and multi-scale lightweight high-resolution human pose estimation network, termed DMLNet, which demonstrates superior performance in comparison to existing lightweight networks.DMLNet introduces the Stem module, TMCneck module, and KTConv module based on HRNet(high-resolution network). The Stem module overcomes the limitations of the original HRNet modules, which could only extract basic information from feature maps, enabling the capture of partial details and deeper information from these maps. The TMCneck module, with its innovative network structure and incorporation of attention mechanisms, significantly enhances the model's accuracy. Meanwhile, the KTConv module achieves both model lightweighting and the extraction of feature information across multiple scales.Additionally, we have introduced dense connections and integrated a feature fusion strategy that combines RFCA attention to enhance the model's performance.We conducted extensive experiments on the COCO and MPII validation datasets, and the results significantly surpass those of existing networks.\u003c/p\u003e","manuscriptTitle":"DMLNet:Densely Connected and Multi-Scale Lightweight High-Resolution Human Pose Estimation Network","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-25 09:21:26","doi":"10.21203/rs.3.rs-7739102/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":"a6a835a4-3f95-4fc0-aa04-91aabcacd6d1","owner":[],"postedDate":"October 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-08T18:09:27+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-25 09:21:26","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7739102","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7739102","identity":"rs-7739102","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.