[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84977-en":3,"doc-seo-84977-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":13,"seo_description":14,"update_tm":28,"read_time":29},84977,7971461740909,"Levi","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Prototype Anchored Generalized Manifold Regression for Unknown Domain Object Detection","Single-Domain Generalized Object Detection (Single-DGOD) transfers a detector trained on one source domain to multiple unseen unknown domains, confronting environmental degradations, localization–recognition coupling, and unobservable distribution gaps. Instead of simulation-driven augmentation with discrete perturbations or static VLM prompts, the work leverages the manifold hypothesis: semantic features lie on a compact low-dimensional manifold. The proposed MR-DCoT uses a Visual–Text dual chain-of-thought to generate structured off-manifold hard examples and a class-specific prototype anchoring module to learn a rectification operator, closing a simulation-to-regression loop that boosts robustness on diverse benchmarks.","Prototype-Anchored Generalized Manifold Regression for Unknown-Domain Object Detection  \nZihao Zhang, Aming Wu, Yang Li and Yahong Han, Member, IEEE.  \narXiv :2607 .07 192v 1 [ cs .CV] 8 Jul 2026  \nAbstract—In this paper, we focus on Single-Domain Generalized Object Detection (Single-DGOD), aiming to transfer a detector trained on one source domain to multiple unknown domains. Existing methods typically rely on simulation-driven paradigms, such as discrete augmentation or static textual prompts, to expand the boundaries of the training distribution. However, finite simulations often fail to capture the infinite dynamic variations of real-world scenarios, which can lead to overfitting on synthetic styles and limit the model’s ability to handle complex structural degradations. Inspired by the manifold hypothesis, we argue that despite diverse visual variations, semantic features inherently reside on a compact, low-dimensional manifold. Thus, the key to generalization lies in learning to rectify deviant samples back onto this stable manifold, rather than merely exhausting external perturbations. To this end, we propose a new framework, i.e., Manifold Regression with Visual-Text Dual Chain-ofThought (MR-DCoT), which reformulates robust generalization as a manifold regression problem. Specifically, we design a VisualText Dual Chain-of-Thought module that couples VLM-guided global semantic evolution with diffusion-based local structural perturbations to generate structured off-manifold hard examples. Subsequently, a Class-Specific Prototype Anchoring mechanism is introduced to learn a robust rectification operator that guides deviant features back toward the source semantic manifold. By establishing a closed loop of simulation for outlier generation and regression for semantic correction, our method effectively bridges the distribution gap, significantly boosting generalization and robustness to unseen shifts. Extensive evaluations on three complementary benchmarks, covering diverse driving weather conditions, real-to-art generalization, and zero-shot semantic segmentation, demonstrate the superiority and versatility of our method in handling complex domain shifts.  \nI. INTRODUCTION  \nSINGLE-domain Generalized Object Detection (Single  \nDGOD) [1]–[3] represents a challenging yet critical task in computer vision. Its objective is to train a detector using only single-source domain data that can robustly adapt to distribution shifts across multiple unseen target domains. However, this task presents significant challenges. First, the simultaneous requirements for localization accuracy and semantic recognition render object detection highly sensitive to environmental degradations, including local texture variations, abrupt illumination changes, and background interference [4],[5] . Second, due to the unobservable distribution gap between the source and unknown domains, traditional feature alignment methods frequently prove ineffective [6], [7] .  \nZihao Zhang, Yang Li, and Yahong Han are with the College of Intelligence and Computing, Tianjin Key Lab of Machine Learning, Tianjin University, Tianjin 300072, China. E-mail: {zhangzihao2490, liyang1389, [yahong](yahong}@tju.edu.cn. Aming Wu)[}](yahong}@tju.edu.cn. Aming Wu)[@tju.edu.cn. Aming Wu](yahong}@tju.edu.cn. Aming Wu) is with the School of Computer Science and Information Engineering, Hefei University of Technology, China. E-mail: [amwu@hfut.edu.cn](amwu@hfut.edu.cn).  \nExisting Single-DGOD methods predominantly follow a Simulation-Driven paradigm: they employ discrete data augmentation [3], style transfer [8], or static textual prompts generated by Vision-Language Models (VLMs) [1], [9], [10] to simulate potential target domains, thereby expanding the boundaries of the training distribution. However, this strategy faces an intrinsic limitation: finite simulation can hardly cover infinite real-world variations. Domain shifts in real-world scenarios (e.g., from clear to stormy weathe","cbCaiumw1Kwi2FHN","https://ap.wps.com/l/cbCaiumw1Kwi2FHN","pdf",16269789,2,1,18,"English","en",105,"# Introduction\n## Single-DGOD challenge and motivation\n## Limitations of simulation-driven methods\n## Manifold hypothesis and regression formulation","[{\"question\":\"What problem does the paper address in object detection?\",\"answer\":\"The paper studies Single-Domain Generalized Object Detection (Single-DGOD), which trains on a single source domain and aims to generalize to multiple unseen target domains with distribution shifts.\"},{\"question\":\"Why do simulation-driven methods have limitations for unknown domains?\",\"answer\":\"Finite simulations cannot cover the infinite, coupled variations of real scenarios, so static perturbations or prompts may overfit to limited synthetic styles and fail under genuine shifts.\"},{\"question\":\"How does the proposed MR-DCoT framework improve generalization?\",\"answer\":\"MR-DCoT reframes robust generalization as manifold regression by generating structured off-manifold hard examples using a Visual–Text dual chain-of-thought module and learning a class-specific prototype anchoring rectification operator to guide deviant features back toward a stable semantic manifold.\"}]",1784199951,45,{"code":4,"msg":31,"data":32},"ok",{"site_id":25,"language":24,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"prototype-anchored-generalized-manifold-regression-for-unknown-domain-object-detection","",{"@graph":36,"@context":85},[37,53,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,47,50],{"item":41,"name":42,"@type":43,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":20},"https://docshare.wps.com/document/","Document",{"item":48,"name":12,"@type":43,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":43,"position":52},"https://docshare.wps.com/document/prototype-anchored-generalized-manifold-regression-for-unknown-domain-object-detection/84977/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":24,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":41,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-24","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the paper address in object detection?","Question",{"text":75,"@type":76},"The paper studies Single-Domain Generalized Object Detection (Single-DGOD), which trains on a single source domain and aims to generalize to multiple unseen target domains with distribution shifts.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why do simulation-driven methods have limitations for unknown domains?",{"text":80,"@type":76},"Finite simulations cannot cover the infinite, coupled variations of real scenarios, so static perturbations or prompts may overfit to limited synthetic styles and fail under genuine shifts.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the proposed MR-DCoT framework improve generalization?",{"text":84,"@type":76},"MR-DCoT reframes robust generalization as manifold regression by generating structured off-manifold hard examples using a Visual–Text dual chain-of-thought module and learning a class-specific prototype anchoring rectification operator to guide deviant features back toward a stable semantic manifold.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":21,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":20,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]