[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82167-en":3,"doc-seo-82167-105":29,"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":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82167,687197207057,"Sage","https://ap-avatar.wpscdn.com/davatar_29158cc5080c5b710cf443261637dec0",8,"Research & Report","STEAM：基于弹性匹配与自适应净化的稳定自训练跨视角地理定位框架","Cross-view geo-localization (CVGL) targets GPS-free localization by matching drone-view images with corresponding satellite-view images. Existing supervised solutions depend on costly, large-scale manually annotated cross-view pairs, while unsupervised methods often rely on generative models or clustering-based stage-wise optimization, causing distribution bias and noisy pseudo-label accumulation. STEAM introduces an end-to-end unsupervised framework that performs self-training directly on real drone and satellite images, using stable feature learning, elastic pseudo-label discovery, and adaptive purification. Experiments on University-1652 and SUES-200 show state-of-the-art performance and comparable results to supervised methods.","STEAM: Stable Self-Training with Elastic Matching and Adaptive Purification  \nShaoxiang Wang 1 , Kejia Zhang 1,* , Haiwei Pan 1 , Lan Zhang2  \n1Harbin Engineering University, School of Computer Science and Technology  \nHarbin, China  \n2Northeast Forestry University, School of Computer and Artificial Intelligence  \nHarbin, China  \n*[kejiazhang@hrbeu.edu.cn](kejiazhang@hrbeu.edu.cn)  \narXiv :2607 .09057v1 [ cs .CV] 10 Jul 2026  \nAbstract  \nCross-view geo-localization (CVGL) aims to achieve GPSfree localization by matching drone-view images with corresponding satellite-view images. Existing supervised methods rely on large-scale manually annotated cross-view image pairs, making them costly and difficult to scale. In contrast, existing unsupervised approaches typically depend on generative models or clustering-based stage-wise optimization, which are prone to distribution bias and the accumulation of noisy pseudo-labels. To address these limitations, we propose STEAM (Stable Self-Training with Elastic Matching and Adaptive Purification), an end-to-end unsupervised cross-view geo-localization framework that performs selftraining directly on real drone and satellite images. Specifically, the proposed Stable Spatial-Aware Module enhances the stability of feature representations, Elastic Matching discovers high-quality cross-view pseudo-labels, and Adaptive Purification dynamically maintains a reliable pseudo-label repository throughout the self-training process. Extensive experiments on the University-1652 and SUES-200 benchmarks demonstrate that STEAM achieves state-of-the-art performance among all existing unsupervised methods and delivers performance comparable to supervised approaches, validating the effectiveness and superiority of the proposed framework. The source code is available at [https://github.com/wsx](https://github.com/wsx)heu/STEAM.git.  \nIntroduction  \nCross-view geo-localization (CVGL) aims to achieve GPSfree localization by matching a query image (e.g., a droneview image) with its corresponding gallery image (e.g., a satellite-view image) in a large-scale image database (Workman, Souvenir, and Jacobs 2015) . As a visual localization technique, CVGL has attracted increasing attention due to its broad applications in autonomous drone navigation, autonomous driving, disaster response, military reconnaissance, and smart cities (Shetty and Gao 2019) . However, the substantial viewpoint discrepancy, resolution inconsistency, and appearance variation between drone and satellite images make this task highly challenging (Deuser, Habel, and Oswald 2023; Wang et al. 2021) .  \nRecent advances in deep learning have significantly promoted the development of cross-view geo-localization. Numerous supervised methods have been proposed, which learn discriminative cross-view feature representations from  \nFigure 1: Comparison between STEAM and existing UCVGL methods. STEAM directly learns from real UAVSatellite data, eliminating fake-image warm-up and clustering stages.  \nmanually annotated drone–satellite image pairs to achieve accurate localization (Durgam et al. 2024; Zhang, Sultani, and Wshah 2023) . However, constructing large-scale paired datasets requires expensive sensing equipment and laborintensive manual annotation, resulting in high data collection costs (Li, Qian, and Xia 2024; Li et al. 2025) . Consequently, learning effective cross-view feature representations without relying on manually paired data has become an important research direction in cross-view geo-localization.  \nTo eliminate the dependency on manual annotations, several unsupervised cross-view geo-localization (UCVGL) methods have recently been proposed, including fake-image warm-up methods (Li, Qian, and Xia 2024) and clusteringbased methods (Wang et al. 2025; Chen et al. 2026) . Although these methods substantially alleviate the reliance on labeled data and achieve promising performance, they still suffer from notable limitations. As illustrated in Fig.","cbCaiukMLngZUSob","https://ap.wps.com/l/cbCaiukMLngZUSob","pdf",2849384,1,9,"English","en",105,"# Abstract\n# Introduction\n## Cross-view geo-localization and challenges\n## Supervised approaches and their limitations\n## Unsupervised methods and remaining issues\n# Proposed Method: STEAM\n## Stable Spatial-Aware Module\n## Elastic Matching\n## Adaptive Purification\n# Related Work","[{\"question\":\"STEAM解决了跨视角地理定位中的哪些关键问题？\",\"answer\":\"STEAM针对现有有监督方法对人工配对数据的依赖，以及无监督方法中生成式建模或聚类阶段带来的分布偏差与噪声伪标签累积问题提出端到端方案。\"},{\"question\":\"STEAM如何在不使用生成图像或聚类初始化的情况下进行自训练？\",\"answer\":\"STEAM直接在真实无人机与卫星图像上进行自训练，避免假图预热、聚类初始化与复杂的分阶段优化。\"},{\"question\":\"STEAM中“弹性匹配”和“自适应净化”分别起什么作用？\",\"answer\":\"弹性匹配通过双向Top-K软匹配与动态阈值过滤发现高质量跨视角伪标签，并兼顾覆盖与正确性；自适应净化通过置信度更新、年龄感知更新与过期标签移除动态维护可靠伪标签库，抑制错误监督的积累。\"}]",1784178548,23,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"steam-stable-self-training-with-elastic-matching-and-adaptive-purification-for-cross-view-geo-localization","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/steam-stable-self-training-with-elastic-matching-and-adaptive-purification-for-cross-view-geo-localization/82167/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","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},"STEAM解决了跨视角地理定位中的哪些关键问题？","Question",{"text":75,"@type":76},"STEAM针对现有有监督方法对人工配对数据的依赖，以及无监督方法中生成式建模或聚类阶段带来的分布偏差与噪声伪标签累积问题提出端到端方案。","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"STEAM如何在不使用生成图像或聚类初始化的情况下进行自训练？",{"text":80,"@type":76},"STEAM直接在真实无人机与卫星图像上进行自训练，避免假图预热、聚类初始化与复杂的分阶段优化。",{"name":82,"@type":73,"acceptedAnswer":83},"STEAM中“弹性匹配”和“自适应净化”分别起什么作用？",{"text":84,"@type":76},"弹性匹配通过双向Top-K软匹配与动态阈值过滤发现高质量跨视角伪标签，并兼顾覆盖与正确性；自适应净化通过置信度更新、年龄感知更新与过期标签移除动态维护可靠伪标签库，抑制错误监督的积累。","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":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":45,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":45,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":45,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":45,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]