[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82240-en":3,"doc-seo-82240-105":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},82240,962075114765,"Quinn","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","HiHR: Hierarchical Hyperbolic Representation for Aerial-Ground Person Re-Identification","Aerial-Ground Person Re-Identification (AG-ReID) retrieves the same individual across heterogeneous aerial and ground cameras, where viewpoint discrepancies cause large appearance shifts that degrade discriminative learning. The HiHR framework introduces multi-granularity features via pre-trained visual-text encoders, then applies Text-guided Multigranularity Fusion (TMF) to strengthen identity representations. Hierarchical Hyperbolic Learning (HHL) builds coarse-to-fine hyperbolic feature structure, preserving cross-view separability and view-specific cues. Experiments on four AG-ReID benchmarks validate effectiveness.","arXiv :2607 .09186v1 [ cs .CV] 10 Jul 2026  \nHiHR: Hierarchical Hyperbolic Representation for Aerial-Ground Person Re-Identification  \nQiwei Yang, Pingping Zhang􀀌  \nDalian University of Technology, Dalian, China  \n􀀌 Corresponding author: zhpp@dlut.edu .cn  \nAbstract. Aerial-Ground Person Re-IDentification (AG-ReID) aims to retrieve the same person across heterogeneous aerial and ground camera platforms. Although great progress has been made, existing methods remain suboptimal due to the direct feature alignment across views, overlooking view-specific cues. To address this issue, we propose a novel Hierarchical Hyperbolic Representation (HiHR) framework for AG-ReID.  \nMore specifically, we first extract multi-granularity features based on pre-trained visual-text encoders. Then, we propose a Text-guided Multigranularity Fusion (TMF) to fuse multi-granularity features and enhance the representation ability of identity features. Furthermore, we introduce the Hierarchical Hyperbolic Learning (HHL) to construct a hierarchical feature structure in a hyperbolic space. This hierarchy includes a coarse level that ensures identity separability and cross-view consistency, and a fine level that preserves view-specific discriminative cues.  \nAs a result, our proposed framework can effectively aggregate viewinvariant and view-specific discriminative features for AG-ReID. Extensive experiments on four AG-ReID benchmarks demonstrate the effectiveness of our framework. The source code is available at [https:](https:)//[github.com/YangQiWei3/HiHR](github.com/YangQiWei3/HiHR).  \nKeywords: Aerial-Ground Person Re-Identification · Hyperbolic Representation Learning · Hierarchical Metric Learning  \n1 Introduction  \nPerson Re-IDentification (ReID) aims to retrieve the same person across nonoverlapping cameras, which serves as a fundamental capability for intelligent surveillance systems. Despite great progress, existing methods [12,21,27] mainly focus on homogeneous camera networks, limiting their applicability to real-world scenarios. In practice, real-world deployments increasingly integrate heterogeneous camera platforms, where ground cameras provide detailed close-range observations, while aerial cameras offer wide-area coverage. This motivates AerialGround Person Re-IDentification (AG-ReID), which aims to retrieve persons across heterogeneous platforms under systematic viewpoint discrepancies. Such discrepancies introduce substantial appearance differences, hindering discriminative representation learning. However, as shown in Fig. 1(a), the discriminative representation should be view-agnostic and allow cross-view discrepancies.  \n2 Q. Yang  \nFig. 1: Illustration of our motivations and proposed framework. (a) The ideal alignment clusters examples of the same identity nearby while different identities far apart. (b) Previous alignments in Euclidean space align cross-view samples directly and may suppress view-specific discriminative cues. (c) Our hierarchical hyperbolic alignment keeps a hierarchical separability and view-specific discriminative cues.  \nTo this end, previous methods [11, 20, 24, 29, 30] directly align samples of the same identity across views, as shown in Fig. 1(b) . Although effective, they emphasize view-invariant features while overlooking view-specific cues. Moreover, many AG-ReID methods [3,5,7,11,33] rely solely on global representations, which capture high-level semantics. They may attenuate mid-level structural cues and fine-grained local patterns for reliable retrieval.  \nTo address the aforementioned issues, we propose a novel Hierarchical Hyperbolic Representation (HiHR) framework for AG-ReID. Specifically, inspired by previous works [12, 21], we first introduce text prompts to extract multigranularity features based on pre-trained visual-text encoders. Then, we propose a Text-guided Multi-granularity Fusion (TMF) to fuse the multi-granularity features and enhance the representation ability of identity features. Furth","cbCaif7Qwl4d2ssk","https://ap.wps.com/l/cbCaif7Qwl4d2ssk","pdf",4910677,2,1,17,"English","en",105,"# Introduction\n# Related Work\n## Aerial-Ground Person Re-Identification","[{\"question\":\"What problem does AG-ReID address?\",\"answer\":\"AG-ReID aims to retrieve the same person across heterogeneous aerial and ground camera platforms under systematic viewpoint discrepancies.\"},{\"question\":\"Why do existing methods perform suboptimally in AG-ReID?\",\"answer\":\"Many approaches directly align features across views or rely mainly on global representations, which can suppress view-specific discriminative cues.\"},{\"question\":\"How does HiHR improve identity representation for cross-view retrieval?\",\"answer\":\"HiHR extracts multi-granularity features using visual-text encoders, fuses them with Text-guided Multigranularity Fusion (TMF), and models coarse-to-fine hierarchical structure in hyperbolic space via Hierarchical Hyperbolic Learning (HHL) to preserve both cross-view consistency and view-specific cues.\"}]",1784179074,43,{"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},"hihr-hierarchical-hyperbolic-representation-for-aerial-ground-person-re-identification","",{"@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/hihr-hierarchical-hyperbolic-representation-for-aerial-ground-person-re-identification/82240/",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","2026-07-19","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 AG-ReID address?","Question",{"text":75,"@type":76},"AG-ReID aims to retrieve the same person across heterogeneous aerial and ground camera platforms under systematic viewpoint discrepancies.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why do existing methods perform suboptimally in AG-ReID?",{"text":80,"@type":76},"Many approaches directly align features across views or rely mainly on global representations, which can suppress view-specific discriminative cues.",{"name":82,"@type":73,"acceptedAnswer":83},"How does HiHR improve identity representation for cross-view retrieval?",{"text":84,"@type":76},"HiHR extracts multi-granularity features using visual-text encoders, fuses them with Text-guided Multigranularity Fusion (TMF), and models coarse-to-fine hierarchical structure in hyperbolic space via Hierarchical Hyperbolic Learning (HHL) to preserve both cross-view consistency and view-specific cues.","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"]