Interactive Text-Guided Image Segmentation via Vision Mamba and Large Language Models

Interactive Text-Guided Image Segmentation via Vision Mamba and Large Language Models | 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 Interactive Text-Guided Image Segmentation via Vision Mamba and Large Language Models Yao Meng, Haochen Sun, Wei Jiang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7965576/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 18 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted 9 You are reading this latest preprint version Abstract We propose an H-type Bidirectional Alignment Network for text-guided image segmentation, achieving efficient and accurate cross-modal feature fusion. The visual branch employs a 12-layer Vision Mamba with four stages, combining a Selective State Space Model (SSM) and local convolutional residual structures to capture long-range dependencies while preserving fine boundaries at reduced computational cost. The text branch, adapted from the Qwen model, freezes lower layers and fine-tunes upper ones to extract robust referential semantics. A Q-Former-based alignment module introduces learnable queries to enforce bidirectional supervision: the forward path performs text-to-image segmentation, and the backward path reconstructs attention from image to text, ensuring cross-modal consistency. A multi-scale decoder integrates aligned features through an iterative interaction mechanism. Experiments on RefCOCO, RefCOCO+, and RefCOCOg demonstrate that our approach outperforms existing methods in both segmentation accuracy and efficiency. Physical sciences/Engineering Physical sciences/Mathematics and computing Text-guided segmentation Vision Mamba Large language models Interactive segmentation Cross-modal alignment Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 18 Mar, 2026 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 31 Dec, 2025 Reviews received at journal 27 Dec, 2025 Reviewers agreed at journal 14 Dec, 2025 Reviews received at journal 13 Dec, 2025 Reviewers agreed at journal 08 Dec, 2025 Reviewers invited by journal 27 Nov, 2025 Editor assigned by journal 26 Nov, 2025 Submission checks completed at journal 13 Nov, 2025 First submitted to journal 13 Nov, 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-7965576","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":552040646,"identity":"9f36654e-76ce-4dd3-94f7-13d7b00066f5","order_by":0,"name":"Yao Meng","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVRIiWNgGAWjYBADOSjNTLwWY9K1JDYQrUW3vffwa54am/Tt7GfMJBgqrBMb2M8ewKvF7My5NGueY2m5O3tygFrOpCc28OQl4NdyI8fMmIftcO6GG7zbJBjbDic2SPAY4Ndy/w1Qy7/D6QZgLf+I0XKDx/gxb9vhBIiWBmK0nMkxY5zbl2a44Uz+Z4uEY+nGbTw5BLQcP2P84c03G3mD48cSb3yosZbtZz+DXwsQsEnxwJgJIC4h9UDA/PEHEapGwSgYBaNgBAMA8QxEeMWhm2gAAAAASUVORK5CYII=","orcid":"","institution":"North China University of Water Resources and Electric Power","correspondingAuthor":true,"prefix":"","firstName":"Yao","middleName":"","lastName":"Meng","suffix":""},{"id":552040647,"identity":"aa525fca-9d41-4a38-87dd-c21173c9c2b6","order_by":1,"name":"Haochen Sun","email":"","orcid":"","institution":"North China University of Water Resources and Electric Power","correspondingAuthor":false,"prefix":"","firstName":"Haochen","middleName":"","lastName":"Sun","suffix":""},{"id":552040648,"identity":"c18e859c-8c02-4eef-86b2-3c9f13ed0079","order_by":2,"name":"Wei Jiang","email":"","orcid":"","institution":"North China University of Water Resources and Electric Power","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Jiang","suffix":""}],"badges":[],"createdAt":"2025-10-28 13:55:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7965576/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7965576/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-026-43841-w","type":"published","date":"2026-03-18T15:59:45+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":97672709,"identity":"74da7207-4994-4c87-8577-aa712a0bc30b","added_by":"auto","created_at":"2025-12-08 09:38:38","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6619386,"visible":true,"origin":"","legend":"","description":"","filename":"paper5.docx","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/fdfdee3b48b0bc1d92b9a7f0.docx"},{"id":97672924,"identity":"ef188d11-9ca4-4867-9e55-f3d89c0ae681","added_by":"auto","created_at":"2025-12-08 09:39:05","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4740,"visible":true,"origin":"","legend":"","description":"","filename":"a5e78b2731404264bf5f03cf07cabeaf.json","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/8a2c834a6d5f6790c5db2188.json"},{"id":97672599,"identity":"e90bde5c-9b72-4bac-b592-a350f3fbb597","added_by":"auto","created_at":"2025-12-08 09:38:29","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":74831,"visible":true,"origin":"","legend":"","description":"","filename":"a5e78b2731404264bf5f03cf07cabeaf1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/058281b3636aab8b4f2cdda6.xml"},{"id":97553680,"identity":"1a31afaf-a8e0-4c65-b356-5a1b82ae771f","added_by":"auto","created_at":"2025-12-05 18:00:37","extension":"png","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155675,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/c1ad39bf12a84e4d88ad8cef.png"},{"id":97670728,"identity":"b55ede0d-b3d8-4571-b393-a05c56a0fe05","added_by":"auto","created_at":"2025-12-08 09:31:12","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1055639,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/e04541e8bed28d365ff299ba.png"},{"id":97672214,"identity":"1601e1cd-70a0-4676-9f58-8b4c500a8bd4","added_by":"auto","created_at":"2025-12-08 09:34:46","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":46146,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/06e607098f68072792ced3c6.png"},{"id":97553686,"identity":"bfd67575-41be-44ea-98e8-7011904b9a36","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25598,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/24e1346bdd38397c677ca83a.png"},{"id":97553685,"identity":"3bb8d0d8-dffa-42bc-80f4-ebee2dc221d4","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":29816,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/bf0f4ec371ed439f99b067ad.png"},{"id":97553687,"identity":"9bc7822a-5158-4068-8810-5cc74949a30f","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":464192,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/2cddf74c2f49321e5a389471.jpeg"},{"id":97670713,"identity":"600bd0b0-0869-4207-acdb-76cd904251f9","added_by":"auto","created_at":"2025-12-08 09:31:12","extension":"wmf","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":688,"visible":true,"origin":"","legend":"","description":"","filename":"image10.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/f180964c190a86f6903c4283.wmf"},{"id":97553690,"identity":"8374f01a-49b1-4412-9261-d14df24b0ecc","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":688,"visible":true,"origin":"","legend":"","description":"","filename":"image11.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/c9f4dded414883bc894af60b.wmf"},{"id":97672729,"identity":"eee0a352-f71d-4bc6-af98-c8abf5c491e7","added_by":"auto","created_at":"2025-12-08 09:38:40","extension":"wmf","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":656,"visible":true,"origin":"","legend":"","description":"","filename":"image12.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/64502dc2cb9ea57af7b95108.wmf"},{"id":97553694,"identity":"c36d0f5c-b959-4b63-b195-eac3537f3768","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":894,"visible":true,"origin":"","legend":"","description":"","filename":"image13.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/5266aec492023e782b114c35.wmf"},{"id":97671343,"identity":"5781af29-3782-4763-8a40-b4f04a245cae","added_by":"auto","created_at":"2025-12-08 09:32:30","extension":"wmf","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":656,"visible":true,"origin":"","legend":"","description":"","filename":"image14.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/7599439024f7ed350a76a6ca.wmf"},{"id":97673155,"identity":"c38cd0c5-311c-4969-b24d-efa640b5c410","added_by":"auto","created_at":"2025-12-08 09:39:31","extension":"wmf","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1730,"visible":true,"origin":"","legend":"","description":"","filename":"image15.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/4ef5703cc1edc3af058c17a8.wmf"},{"id":97673051,"identity":"c12feb14-1754-4d47-8dd9-2ac14fdfb58f","added_by":"auto","created_at":"2025-12-08 09:39:20","extension":"wmf","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":786,"visible":true,"origin":"","legend":"","description":"","filename":"image16.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/d63059ad396b83f3fa8385e7.wmf"},{"id":97672282,"identity":"4332523b-3634-4568-aa43-cad5d2c4527e","added_by":"auto","created_at":"2025-12-08 09:35:03","extension":"wmf","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":784,"visible":true,"origin":"","legend":"","description":"","filename":"image17.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/df2e9273620f4c8a6ba0f09b.wmf"},{"id":97553695,"identity":"30a1df45-5bc7-4383-92d5-ff8c3a814542","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":898,"visible":true,"origin":"","legend":"","description":"","filename":"image18.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/860fb6c98513e393e2d31cfd.wmf"},{"id":97553698,"identity":"cdb565d0-6ce2-4224-9359-90aa992069ea","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":3258,"visible":true,"origin":"","legend":"","description":"","filename":"image19.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/88f36d36a7a2343916055a88.wmf"},{"id":97553699,"identity":"6513d2fd-65de-48e3-97ad-0621a973471b","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1044,"visible":true,"origin":"","legend":"","description":"","filename":"image20.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/4aa57e0b1e81976c8b2887ff.wmf"},{"id":97672760,"identity":"781a12a3-6dbe-4190-ae68-6b29956e8c84","added_by":"auto","created_at":"2025-12-08 09:38:45","extension":"wmf","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":656,"visible":true,"origin":"","legend":"","description":"","filename":"image21.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/3a0320dbeb8d7434a25dfb3e.wmf"},{"id":97553719,"identity":"c942974f-c7fb-4e8b-b3e4-80843e4b5595","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"wmf","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":946,"visible":true,"origin":"","legend":"","description":"","filename":"image22.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/3461651d28bea764c2792b5f.wmf"},{"id":97553707,"identity":"bcf802d3-23a7-4e67-8cc5-b707fd2ab1cb","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2190,"visible":true,"origin":"","legend":"","description":"","filename":"image25.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/0d76d2f43abc0403ded30145.wmf"},{"id":97672299,"identity":"3f3ccee4-fcb6-48a2-a8c8-c302c0556a42","added_by":"auto","created_at":"2025-12-08 09:35:15","extension":"wmf","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1422,"visible":true,"origin":"","legend":"","description":"","filename":"image26.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/12c181188b57a330038c80a7.wmf"},{"id":97672716,"identity":"26e53960-641d-40c7-82dc-dce97118afe2","added_by":"auto","created_at":"2025-12-08 09:38:38","extension":"wmf","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2636,"visible":true,"origin":"","legend":"","description":"","filename":"image27.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/1eef04a878ceb8ce2ab4b831.wmf"},{"id":97553701,"identity":"c5792bef-f9ff-42d7-939a-ab4493b16081","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":442,"visible":true,"origin":"","legend":"","description":"","filename":"image28.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/147f9cfc7c3625e0bd5c6cdb.wmf"},{"id":97553703,"identity":"de5086f0-58e5-460f-9214-18a108c980b7","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1298,"visible":true,"origin":"","legend":"","description":"","filename":"image29.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/aafefc6bec2e130d9b9378f4.wmf"},{"id":97553700,"identity":"062a3b08-6506-48a0-aeba-94e59b4b055a","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"wmf","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":678,"visible":true,"origin":"","legend":"","description":"","filename":"image9.wmf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/e71aa83da8a3f63025944096.wmf"},{"id":97553708,"identity":"98772b6c-12fe-4754-9003-8c071c6bdc9f","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42951,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/0e72de08e9ed5d69af7e32ba.png"},{"id":97553724,"identity":"4f400fbb-054c-4656-bca3-25860efd171e","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":179231,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/0081304729b3fdd95dcc7705.png"},{"id":97672693,"identity":"88b01ae3-d7c7-40d9-8498-8b483bddebed","added_by":"auto","created_at":"2025-12-08 09:38:35","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":21785,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/83ac5ddd4705baa19ca12262.png"},{"id":97672404,"identity":"13e4fee7-6f16-40dc-8123-83b045f0d5dc","added_by":"auto","created_at":"2025-12-08 09:37:17","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":10859,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/2eec8a1ee3a8f79912b52a72.png"},{"id":97553704,"identity":"e0239d5b-05e4-4c43-86a5-748efb975fde","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":14076,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/9fba26fc108866f186c23d4b.png"},{"id":97672884,"identity":"e315da9b-7211-4d70-bd58-a537bef8a1da","added_by":"auto","created_at":"2025-12-08 09:38:59","extension":"png","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":152782,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/e8cdfcc00e6b22c2d754fe1c.png"},{"id":97553727,"identity":"66a1eb42-a44b-4c90-a84d-4ebfe4ba1956","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":34,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":415,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/f5763574e19eb31ce9aedc01.png"},{"id":97553728,"identity":"ab527eee-83ad-4617-901b-fcea2f8b290c","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":35,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":336,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/221839cb1cd52a40e984d257.png"},{"id":97553726,"identity":"12dbd66c-ed8d-4d02-9521-29eb54788f7f","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":36,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":230,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage12.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/e40f20ce011fe9f381403a8f.png"},{"id":97553702,"identity":"9b7b3b61-1d52-43de-a166-891385fa06b9","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":37,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":368,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage13.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/36a5572d642875ba72b42064.png"},{"id":97672667,"identity":"e4540b58-704e-403a-8e64-6db8af9ce48a","added_by":"auto","created_at":"2025-12-08 09:38:32","extension":"png","order_by":38,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":255,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage14.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/2281874f23ec4b822dd9055d.png"},{"id":97671359,"identity":"a2124159-6f6a-4cde-b408-e41484ac6e2d","added_by":"auto","created_at":"2025-12-08 09:32:31","extension":"png","order_by":39,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1377,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage15.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/4cb86e3f507ba79492554ce0.png"},{"id":97553720,"identity":"92109f65-d8ad-4aa3-8ba6-7780aef7fdde","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":40,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":404,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage16.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/871e41482445369190d19b8b.png"},{"id":97553735,"identity":"e7c86b21-892c-4317-ba96-848fdd1f26ff","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":41,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":303,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage17.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/acb812cbba880950db4880ff.png"},{"id":97673066,"identity":"a1ca78b7-fd6a-4aae-aca5-cedfd216693b","added_by":"auto","created_at":"2025-12-08 09:39:23","extension":"png","order_by":42,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":372,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage18.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/51ae6286d3d4c8e0a902115e.png"},{"id":97553723,"identity":"a3f00fe3-908c-4692-a7fa-eb469aacbc2f","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":43,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1979,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage19.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/c76822004694f96c725c6a61.png"},{"id":97553732,"identity":"dd6b9a2b-78e3-47b5-a9e4-6e547d41f4cc","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":44,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":609,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage20.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/1c22230313cfb434ccb176c5.png"},{"id":97553709,"identity":"22abcbeb-973d-4f88-abff-1f426e6d686a","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":45,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":219,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage21.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/8acef4d600ddf047b4b4046c.png"},{"id":97553710,"identity":"f763e034-6ac0-4af6-860b-5bf6dc8f3769","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":46,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":490,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage22.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/bfdc7e95472800ff77940ea2.png"},{"id":97553730,"identity":"0dff1581-f9e1-46cd-9eac-702508bb7a84","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":47,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1879,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage25.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/951c6a799d8dacec07ddc7ca.png"},{"id":97672561,"identity":"d57979c5-a742-43a0-a503-b378eea6ed84","added_by":"auto","created_at":"2025-12-08 09:38:23","extension":"png","order_by":48,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1606,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage26.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/4c82816167a5d7e5c0983971.png"},{"id":97553711,"identity":"17196417-5f4f-419b-bd7d-21bab2c45fc6","added_by":"auto","created_at":"2025-12-05 18:00:38","extension":"png","order_by":49,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2965,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage27.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/b6ea184cc607b83f3b7158a7.png"},{"id":97672750,"identity":"2ed57d49-de67-4977-94e7-d1f8d59e60d7","added_by":"auto","created_at":"2025-12-08 09:38:44","extension":"png","order_by":50,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":132,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage28.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/168c68e31cae117235d5924b.png"},{"id":97672718,"identity":"d424d382-9746-48c6-af56-c66c8a1b7012","added_by":"auto","created_at":"2025-12-08 09:38:38","extension":"png","order_by":51,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1013,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage29.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/b515d9e2144b35218c0a9167.png"},{"id":97553725,"identity":"31c6997a-378d-40a9-bac3-55160681f890","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"png","order_by":52,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":294,"visible":true,"origin":"","legend":"","description":"","filename":"Onlineimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/27537063676220cd680ea033.png"},{"id":97673270,"identity":"994c80a2-bf10-479b-a27d-d7c1f342927e","added_by":"auto","created_at":"2025-12-08 09:39:46","extension":"xml","order_by":53,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":73983,"visible":true,"origin":"","legend":"","description":"","filename":"a5e78b2731404264bf5f03cf07cabeaf1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/0032806aa472fd53de9bb686.xml"},{"id":97553721,"identity":"8eae77c3-f1c5-4326-8756-73fd45390c5e","added_by":"auto","created_at":"2025-12-05 18:00:39","extension":"html","order_by":54,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":80852,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1/e6a649f96c5b67cfc52fd86a.html"},{"id":105223747,"identity":"796883b1-7bd0-4823-b3dd-eaafc38dc09f","added_by":"auto","created_at":"2026-03-23 16:09:49","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":587511,"visible":true,"origin":"","legend":"","description":"","filename":"paper5.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7965576/v1_covered_42fc862e-0936-4df7-bdaf-5f81b86790ec.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Interactive Text-Guided Image Segmentation via Vision Mamba and Large Language Models","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"Text-guided segmentation, Vision Mamba, Large language models, Interactive segmentation, Cross-modal alignment","lastPublishedDoi":"10.21203/rs.3.rs-7965576/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7965576/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWe propose an H-type Bidirectional Alignment Network for text-guided image segmentation, achieving efficient and accurate cross-modal feature fusion. The visual branch employs a 12-layer Vision Mamba with four stages, combining a Selective State Space Model (SSM) and local convolutional residual structures to capture long-range dependencies while preserving fine boundaries at reduced computational cost. The text branch, adapted from the Qwen model, freezes lower layers and fine-tunes upper ones to extract robust referential semantics. A Q-Former-based alignment module introduces learnable queries to enforce bidirectional supervision: the forward path performs text-to-image segmentation, and the backward path reconstructs attention from image to text, ensuring cross-modal consistency. A multi-scale decoder integrates aligned features through an iterative interaction mechanism. Experiments on RefCOCO, RefCOCO+, and RefCOCOg demonstrate that our approach outperforms existing methods in both segmentation accuracy and efficiency.\u003c/p\u003e","manuscriptTitle":"Interactive Text-Guided Image Segmentation via Vision Mamba and Large Language Models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-05 18:00:33","doi":"10.21203/rs.3.rs-7965576/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-12-31T17:18:11+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-27T07:57:05+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"158230127003224078580551809823568095625","date":"2025-12-15T02:12:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-12-13T15:18:42+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"14461457135168045246477989274450791783","date":"2025-12-08T13:22:25+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-11-27T11:48:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-11-26T08:31:49+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-11-14T02:20:34+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-11-14T02:17:37+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":"3b3c950d-fdbc-43c7-9636-a52151b34616","owner":[],"postedDate":"December 5th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":58745746,"name":"Physical sciences/Engineering"},{"id":58745747,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2026-03-23T16:06:55+00:00","versionOfRecord":{"articleIdentity":"rs-7965576","link":"https://doi.org/10.1038/s41598-026-43841-w","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2026-03-18 15:59:45","publishedOnDateReadable":"March 18th, 2026"},"versionCreatedAt":"2025-12-05 18:00:33","video":"","vorDoi":"10.1038/s41598-026-43841-w","vorDoiUrl":"https://doi.org/10.1038/s41598-026-43841-w","workflowStages":[]},"version":"v1","identity":"rs-7965576","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7965576","identity":"rs-7965576","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}