[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2165":3},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":23,"tags":25,"view_count":31,"doi":32,"paper":33,"created_at":62},2165,"Multi-grained image-text retrieval in farmland remote sensing via multimodal large language models","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.engappai.2026.116113","Low-altitude farmland remote sensing (FRS) has become an important tool for precision agricultural monitoring. However, existing FRS methods are mainly designed for segmentation with predefined categories, making them insufficient for semantic understanding and unable to support flexible natural-language queries. To address this limitation, we introduce an artificial intelligence-based task, FRS cross-modal retrieval, which aims to retrieve target farmland regions from large-scale remote sensing images using textual descriptions. To support this task, we construct the first Farmland Remote Sensing Cross-Modal Retrieval (FRS-CMR) dataset, containing 67,806 image-text pairs generated through multimodal large language models and expert refinement. Based on FRS-CMR, we propose a Farmland-oriented Multi-Grained Retrieval (FarmMGR) framework for image-text alignment. FarmMGR jointly exploits global scene semantics and local region-level cues to capture both large-scale landscape patterns and fine-grained farmland characteristics. To further improve semantic alignment, we design Prototype-Based Semantic Alignment (PBSA) to alleviate the soft-positive sample problem, Global Semantics Hierarchical Feature Integration (GSHFI) to adaptively fuse global and local representations, and Instance-Level Semantic Relation Modeling (ISRM) to capture relational dependencies among farmland instances. In addition, Memory-Enhanced Cross-Modal Contrastive Learning (MECCL) enlarges the negative sample pool and improves representation discrimination. Extensive experiments on FRS-CMR show that FarmMGR achieves a mean recall of 28.57%, outperforming existing image-text retrieval methods and demonstrating its effectiveness for language-driven farmland remote sensing retrieval.","低空农田遥感(FRS)已成为精准农业监测的重要工具。然而，现有农田遥感方法主要针对预定义类别的分割任务设计，难以进行语义理解，也无法支持灵活的自然语言查询。为解决这一局限，我们引入了一项基于人工智能的任务——农田遥感跨模态检索，旨在利用文本描述从大规模遥感图像中检索目标农田区域。为支撑该任务，我们构建了首个农田遥感跨模态检索(FRS-CMR)数据集，包含67,806个通过多模态大语言模型生成并经专家精炼的图像-文本对。基于FRS-CMR，我们提出了面向农田的多粒度检索(FarmMGR)框架以实现图像-文本对齐。FarmMGR联合利用全局场景语义和局部区域级线索，以同时捕获大尺度景观格局和细粒度农田特征。为进一步提升语义对齐效果，我们设计了基于原型的语义对齐(PBSA)以缓解软正样本问题，全局语义层次特征融合(GSHFI)以自适应融合全局与局部表示，以及实例级语义关系建模(ISRM)以捕获农田实例间的关系依赖。此外，记忆增强跨模态对比学习(MECCL)扩大了负样本池并提升了表示判别能力。在FRS-CMR上的大量实验表明，FarmMGR取得了28.57%的平均召回率，优于现有图像-文本检索方法，证明了其在语言驱动农田遥感检索中的有效性。",null,"Engineering Applications of Artificial Intelligence","2026-09-10T00:00:00Z","论文",10,false,81,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,14,9,1,"构建首个农田遥感跨模态检索数据集并提出多粒度对齐框架，方法新颖、数据规模可观，对语言驱动的精准农业监测有实质推动。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","农业人工智能","精准农业","遥感","多模态大模型",0,"10.1016\u002Fj.engappai.2026.116113",{"doi":32,"openalex_id":34,"authors":35,"venue":10,"cited_by_count":31,"oa_url":6,"card":55,"direction":59,"ingested_from":61},"W7212180335",[36,38,41,43,46,48,50,52],{"name":37,"orcid":9},"KeJian Yu",{"name":39,"orcid":40},"Wentao Ma","https:\u002F\u002Forcid.org\u002F0000-0002-2781-1693",{"name":42,"orcid":9},"Shichao Jin",{"name":44,"orcid":45},"Lu Liu","https:\u002F\u002Forcid.org\u002F0000-0002-3170-9376",{"name":47,"orcid":9},"Yuwei Wang",{"name":49,"orcid":9},"Weiwei Wang",{"name":51,"orcid":9},"Longzhe Quan",{"name":53,"orcid":54},"Lichuan Gu","https:\u002F\u002Forcid.org\u002F0000-0002-3768-8203",{"tldr":56,"method":57,"finding":58,"direction":59,"opportunity":60},"提出农田遥感图文跨模态检索任务，构建数据集并提出FarmMGR框架实现语言驱动检索。","构建67,806图文对FRS-CMR数据集，用多粒度对齐与记忆增强对比学习。","FarmMGR平均召回率28.57%，优于现有图文检索方法，验证语言驱动农田遥感检索可行。","农业遥感与作物表型","可探索多模态大模型生成数据的噪声校正、跨区域泛化及细粒度农田语义检索基准。","openalex","2026-09-11T23:30:30.005240Z"]