[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3046":3,"related-3046":80},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":8,"source_name":9,"source_url":8,"published_at":10,"category":11,"cover_url":8,"hotness":12,"is_selected":13,"score":14,"score_detail":15,"sources":22,"tags":27,"primary_topic":33,"secondary_topics":39,"classification":40,"search_phrases":64,"slug":67,"view_count":68,"doi":8,"paper":69,"created_at":79},3046,"基于Flor-YOLO的香石竹鲜切花分级轻量化检测方法","https:\u002F\u002Ffinance.sina.com.cn\u002Froll\u002F2026-09-20\u002Fdoc-inisnssw6528969.shtml","西南林业大学李传孟、杨洁副教授、张晓宇在《智慧农业（中英文）》2026,8(4):85-99发表Flor-YOLO模型，针对香石竹鲜切花开放度人工分级主观性强、效率低问题，以YOLO11n为基线进行骨干网络、下采样方式、检测头结构针对性改进。Flor-YOLO在自建香石竹数据集上mAP@50达到96.10%，较基准模型提升3.25个百分点；模型参数量与计算量分别为1.26M和1.1GFLOPs，同比降低51.2%和82.5%。",null,"智慧农业(中英文)2026,8(4):85-99","2026-09-20T12:26:00Z","论文",25,false,83,{"impact":16,"substance":17,"depth":16,"authority":18,"freshness":19,"relevant":20,"comment":21},18,23,14,10,1,"核心期刊论文，提出轻量化Flor-YOLO模型实现香石竹鲜切花自动分级，数据翔实、方法有创新，对花卉产业智能化有参考价值。",[23,24],{"name":9,"url":6},{"name":25,"url":26},"《智慧农业（中英文）》2026年第4期","https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7687597315586818611",[28,29,30,31,32],"智慧农业","农业人工智能","花卉产业","轻量化模型","鲜切花分级",{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":41,"locked":13,"version":42,"method":43,"evidence":44},"classified","2026-10-22.2","rules",{"candidates":45,"agricultural_evidence":48},[46,53,59],{"code":34,"score":47,"supported":48,"title":49,"abstract":51,"tags":52},13,true,[50],"YOLO",[50],[29,31],{"code":54,"score":55,"supported":48,"title":56,"abstract":57,"tags":58},"I1",4,[],[28],[28],{"code":60,"score":20,"supported":13,"title":61,"abstract":62,"tags":63},"B4",[],[],[30],[65,66],"西南林业大学 香石竹 鲜切花分级","Flor-YOLO 香石竹 检测","西南林业大学香石竹鲜切花分级-3046",3,{"doi":8,"openalex_id":8,"authors":70,"venue":8,"cited_by_count":71,"oa_url":8,"card":72,"direction":76,"ingested_from":78},[],0,{"tldr":73,"method":74,"finding":75,"direction":76,"opportunity":77},"提出轻量化Flor-YOLO模型，实现香石竹鲜切花开放度自动分级检测。","以YOLO11n为基线，改进骨干网络、下采样方式与检测头，自建香石竹数据集。","mAP@50达96.10%，较基准提升3.25个百分点，参数量与计算量分别降低51.2%和82.5%","农业人工智能与决策模型","可迁移至其他花卉或果蔬的轻量化分级，并探索边缘设备实时部署与多任务联合检测。","agent","2026-09-21T00:04:39.154232Z",{"total":81,"page":20,"page_size":81,"items":82},6,[83,136,182,266,335,399],{"id":84,"title":85,"url":86,"summary":87,"summary_zh":88,"content":8,"source_name":89,"source_url":86,"published_at":90,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":91,"score_detail":92,"sources":97,"tags":99,"primary_topic":102,"secondary_topics":103,"classification":104,"search_phrases":116,"slug":119,"view_count":68,"doi":120,"paper":121,"created_at":135},3517,"A Resource-Efficient Hybrid CNN-LSTM Network for Image-Based Bean Leaf Disease Classification","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fjimaging12100468","Accurate and resource-efficient automated diagnosis is a cornerstone of modern agricultural expert systems. While Convolutional Neural Networks (CNNs) have established benchmarks in plant pathology, their ability to capture long-range spatial dependencies is often limited by standard pooling layers, and their high memory footprint hinders deployment on portable devices. This paper proposes a lightweight hybrid CNN-LSTM system for bean leaf disease classification. By integrating an LSTM layer to model the spatial–sequential relationships within feature maps, our hybrid architecture achieves a 94.36% accuracy and 94.38% F1 score while maintaining an exceptionally small footprint of 1.86 MB, a 70% reduction in size compared to traditional CNN-based systems. Furthermore, we provide a systematic evaluation of image augmentation strategies, demonstrating that tailored transformations are superior to generic combinations for maintaining the integrity of diagnostic patterns. Results on the ibean dataset confirm that the proposed system achieves state-of-the-art F1 scores of 99.22% with EfficientNet-B7+LSTM, providing a potentially robust and scalable framework for real-time agricultural decision support in resource-constrained environments. The code and augmented datasets used in this study are publicly available on this GitHub repo.","准确且资源高效的自动化诊断是现代农业专家系统的基石。尽管卷积神经网络（CNN）在植物病理学领域已确立了基准，但其捕捉长程空间依赖关系的能力常受限于标准池化层，且高内存占用阻碍了其在便携设备上的部署。本文提出了一种用于豆叶病害分类的轻量级混合CNN-LSTM系统。通过集成LSTM层来建模特征图内的空间-序列关系，我们的混合架构达到了94.36%的准确率和94.38%的F1分数，同时保持了仅1.86 MB的极小占用，相较于传统基于CNN的系统体积减少了70%。此外，我们系统评估了图像增强策略，表明定制化变换在保持诊断模式完整性方面优于通用组合。在ibean数据集上的结果证实，所提出的系统结合EfficientNet-B7+LSTM达到了99.22%的最先进F1分数，为资源受限环境中的实时农业决策支持提供了一个潜在稳健且可扩展的框架。本研究使用的代码和增强数据集已在此GitHub仓库公开。","Journal of Imaging","2026-09-24T00:00:00Z",78,{"impact":93,"substance":94,"depth":16,"authority":47,"freshness":95,"relevant":20,"comment":96},16,22,9,"轻量级CNN-LSTM豆叶病害识别研究，方法新颖、数据可靠且代码开源，对资源受限场景下的农业智能诊断有实用价值。",[98],{"name":89,"url":86},[28,29,100,31,101],"病害识别","豆类作物",{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},[],{"status":41,"locked":13,"version":42,"method":43,"evidence":105},{"candidates":106,"agricultural_evidence":48},[107,112],{"code":34,"score":81,"supported":48,"title":108,"abstract":109,"tags":111},[],[110],"decision support",[29,100,31],{"code":54,"score":20,"supported":13,"title":113,"abstract":114,"tags":115},[],[],[28],[117,118],"ibean dataset 豆叶病害","农业人工智能 轻量化模型 智慧农业 病害识别","ibeandataset豆叶病害-3517","10.3390\u002Fjimaging12100468",{"doi":120,"openalex_id":122,"authors":123,"venue":89,"cited_by_count":71,"oa_url":86,"card":129,"direction":76,"ingested_from":134},"W7154572440",[124,126],{"name":125,"orcid":8},"Hye Jin Rhee",{"name":127,"orcid":128},"Joseph Damilola Akinyemi","https:\u002F\u002Forcid.org\u002F0000-0003-3121-4231",{"tldr":130,"method":131,"finding":132,"direction":76,"opportunity":133},"提出轻量级CNN-LSTM混合网络，用于豆叶病害分类，兼顾高精度与低资源占用。","CNN提取特征后接LSTM建模空间序列关系，在ibean数据集上评估并系统比较图","模型准确率94.36%、F1 94.38%，仅1.86MB，比传统CNN缩小70%，Efficien","可探索面向移动端\u002F边缘设备的超轻量病害诊断模型，并研究增强策略与模型结构的自适应协同优化。","openalex","2026-09-25T23:30:59.276381Z",{"id":137,"title":138,"url":139,"summary":140,"summary_zh":8,"content":8,"source_name":141,"source_url":8,"published_at":142,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":143,"score_detail":144,"sources":146,"tags":148,"primary_topic":151,"secondary_topics":152,"classification":153,"search_phrases":170,"slug":173,"view_count":81,"doi":8,"paper":174,"created_at":181},3048,"基于双路径注意力与多尺度融合的作物病虫害识别网络DPMFNet","https:\u002F\u002Fwww.mdpi.com\u002F1099-4300\u002F28\u002F9\u002F1032","盐城工学院Hong Zhang、Fagen Song等联合江苏开放大学提出DPMFNet轻量级双路径网络，集成空间-通道双注意力（SCDA）与多尺度深度可分离卷积（MDSC）模块，构建AttMDSCBlock残差结构。在PlantVillage与AI Challenger 2018数据集上DPMFNet仅14.24M参数和2.55G FLOPs，跨注意力机制融合局部细节与全局上下文，轻量化金字塔策略自适应整合多分辨率特征，在复杂农田场景下兼顾精度与可部署性，为嵌入式田间设备提供高性价比方案。","MDPI Entropy 28(9):1032","2026-09-20T00:00:00Z",79,{"impact":93,"substance":94,"depth":16,"authority":47,"freshness":19,"relevant":20,"comment":145},"轻量级双路径注意力网络在两大公开数据集上兼顾精度与可部署性，对嵌入式田间设备落地有实质参考价值。",[147],{"name":141,"url":139},[28,29,149,150,31],"病虫害识别","作物监测",{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},[],{"status":41,"locked":13,"version":42,"method":43,"evidence":154},{"candidates":155,"agricultural_evidence":48},[156,161,165],{"code":34,"score":157,"supported":48,"title":158,"abstract":159,"tags":160},11,[149],[],[29,149,31],{"code":54,"score":20,"supported":13,"title":162,"abstract":163,"tags":164},[],[],[28],{"code":166,"score":20,"supported":13,"title":167,"abstract":168,"tags":169},"I3",[],[],[150],[171,172],"盐城工学院 DPMFNet 病虫害识别","PlantVillage AI Challenger 作物病害","盐城工学院DPMFNet病虫害识别-3048",{"doi":8,"openalex_id":8,"authors":175,"venue":8,"cited_by_count":71,"oa_url":8,"card":176,"direction":76,"ingested_from":78},[],{"tldr":177,"method":178,"finding":179,"direction":76,"opportunity":180},"提出轻量级双路径网络DPMFNet，实现复杂农田场景下的作物病虫害高精度识别。","空间-通道双注意力与多尺度深度可分离卷积，构建AttMDSCBlock残差结构。","仅14.24M参数、2.55G FLOPs，在PlantVillage与AI Challenger ","可探索真实田间多病虫害并发与边缘设备实时推理的轻量化自适应识别研究。","2026-09-21T00:04:39.305757Z",{"id":183,"title":184,"url":185,"summary":186,"summary_zh":187,"content":8,"source_name":188,"source_url":185,"published_at":189,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":190,"score_detail":191,"sources":193,"tags":195,"primary_topic":198,"secondary_topics":204,"classification":211,"search_phrases":237,"slug":240,"view_count":241,"doi":242,"paper":243,"created_at":265},2646,"IoT-RiceMobileNet: An improved lightweight MobileNetV2 Model for real-time multi-class rice disease detection using IoT","https:\u002F\u002Fdoi.org\u002F10.1371\u002Fjournal.pone.0356383","Early and real-time detection of rice leaf diseases (RLD) poses a significant challenge for farmers, especially in rural regions with limited access to advanced technology. Conventional deep learning models often require substantial computational resources, rendering them impractical for deployment on mobile or edge devices commonly used in agricultural environments. Although models such as ResNet50, VGG16, and InceptionV3 can achieve high accuracy, they are computationally expensive and may be less suitable for real-time deployment in smart agricultural systems. Furthermore, many existing studies rely on limited datasets, which can restrict model generalizability under diverse real-world conditions. Although transfer learning can improve classification performance, developing lightweight models suitable for real-time IoT deployment remains challenging. To address these limitations, we curated a hybrid dataset of 7,092 rice leaf images by combining self-collected and Kaggle samples and proposed IoT-RiceMobileNet, a lightweight improved MobileNetV2-based model that can be effectively integrated into our developed IoT system. The proposed model outperformed transfer learning and deep learning baseline models, achieving 99.19% test accuracy and 99.20% precision while maintaining low computational complexity. We further confirmed model stability using stratified 5-fold and 10-fold cross-validation on both the constructed RLD and multi-source datasets, achieving mean accuracies of 98.04% and 98.13% on the constructed RLD dataset and 98.11% and 98.58% on the multi-source dataset, respectively. Moreover, our models compact size and 85.17 FPS inference speed support real-time deployment. Finally, the model was integrated into an IoT-enabled mobile, web, and cloud-based inference framework, demonstrating its practical potential for scalable rice disease detection in smart agriculture.","水稻叶片病害（RLD）的早期和实时检测对农民而言是一项重大挑战，尤其是在难以获取先进技术的农村地区。传统的深度学习模型通常需要大量计算资源，使其难以在实际农业环境中常用的移动或边缘设备上部署。尽管ResNet50、VGG16和InceptionV3等模型可以达到较高精度，但其计算成本高昂，可能不太适合在智慧农业系统中实时部署。此外，许多现有研究依赖有限的数据集，这可能限制模型在多样化真实条件下的泛化能力。尽管迁移学习可以提升分类性能，但开发适合实时物联网部署的轻量级模型仍具有挑战性。为解决这些局限，我们通过整合自采集样本和Kaggle样本，构建了一个包含7092张水稻叶片图像的混合数据集，并提出了IoT-RiceMobileNet，这是一种基于MobileNetV2改进的轻量级模型，可有效集成到我们开发的物联网系统中。所提出的模型优于迁移学习和深度学习基线模型，在保持低计算复杂度的同时，实现了99.19%的测试准确率和99.20%的精确率。我们进一步通过在构建的RLD数据集和多源数据集上采用分层5折和10折交叉验证确认了模型稳定性，在构建的RLD数据集上分别达到98.04%和98.13%的平均准确率，在多源数据集上分别达到98.11%和98.58%的平均准确率。此外，该模型紧凑的规模和85.17 FPS的推理速度支持实时部署。最后，该模型被集成到一个支持物联网的移动端、网页端和云端推理框架中，展示了其在智慧农业中可扩展水稻病害检测的实际潜力。","PLoS ONE","2026-09-15T00:00:00Z",80,{"impact":16,"substance":94,"depth":16,"authority":47,"freshness":95,"relevant":20,"comment":192},"提出轻量级MobileNetV2水稻病害检测模型并集成物联网推理框架，实测精度与推理速度俱佳，方法新颖、数据规模可观，对边缘端智慧农业落地有参考价值。",[194],{"name":188,"url":185},[28,29,196,197,31],"物联网","水稻病害",{"code":199,"slug":200,"name":201,"path":202,"parent_code":203},"A1","rice","水稻","\u002Ftopic\u002Ffield-crops\u002Frice","A",[205,210],{"code":206,"slug":207,"name":208,"path":209,"parent_code":38},"I5","agricultural-iot","物联网传感与边缘计算","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-iot",{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},{"status":41,"locked":13,"version":42,"method":43,"evidence":212},{"candidates":213,"agricultural_evidence":48},[214,219,224,231],{"code":199,"score":215,"supported":48,"title":216,"abstract":217,"tags":218},15,[200],[201,200],[197],{"code":206,"score":215,"supported":48,"title":220,"abstract":222,"tags":223},[221],"IoT",[196,221],[196],{"code":34,"score":225,"supported":48,"title":226,"abstract":227,"tags":230},8,[],[228,229],"深度学习","deep learning",[29,31],{"code":54,"score":232,"supported":48,"title":233,"abstract":234,"tags":236},7,[],[28,235],"smart agriculture",[28],[238,239],"农业人工智能 轻量化模型 智慧农业 水稻病害","农业人工智能 轻量化模型","农业人工智能轻量化模型智慧农业水稻病害-2646",2,"10.1371\u002Fjournal.pone.0356383",{"doi":242,"openalex_id":244,"authors":245,"venue":188,"cited_by_count":71,"oa_url":185,"card":259,"direction":263,"ingested_from":134},"W7213328321",[246,248,250,252,254,256],{"name":247,"orcid":8},"Khawja Imran Masud",{"name":249,"orcid":8},"Md Yasin Zihad",{"name":251,"orcid":8},"Mehedi Hasan Shuvo",{"name":253,"orcid":8},"Mst Raonik Jannat",{"name":255,"orcid":8},"Jia Uddin",{"name":257,"orcid":258},"Sahara Ali","https:\u002F\u002Forcid.org\u002F0000-0002-8578-948X",{"tldr":260,"method":261,"finding":262,"direction":263,"opportunity":264},"提出轻量级IoT-RiceMobileNet模型，实现水稻病害实时多类检测。","改进MobileNetV2，融合自采与Kaggle共7092张图像，IoT端部署","测试准确率99.19%，推理速度85.17 FPS，适合边缘设备实时检测。","智慧农业 \u002F 农业物联网","可探索多作物病害泛化、田间复杂光照下轻量模型鲁棒性及边缘联邦学习。","2026-09-16T23:30:12.197386Z",{"id":267,"title":268,"url":269,"summary":270,"summary_zh":271,"content":8,"source_name":272,"source_url":269,"published_at":273,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":274,"score_detail":275,"sources":279,"tags":281,"primary_topic":284,"secondary_topics":285,"classification":286,"search_phrases":301,"slug":304,"view_count":55,"doi":305,"paper":306,"created_at":334},2285,"PetalSpot: an open-source two-stage deep-learning workflow for pixel-level quantification of petal lesions","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.compag.2026.112421","PetalSpot: an open-source two-stage deep-learning workflow for pixel-level quantification of petal lesions。Computers and Electronics in Agriculture","PetalSpot：一种用于花瓣病斑像素级量化的开源两阶段深度学习工作流程。","Computers and Electronics in Agriculture","2026-09-11T00:00:00Z",75,{"impact":93,"substance":276,"depth":277,"authority":18,"freshness":225,"relevant":20,"comment":278},20,17,"开源两阶段深度学习流程实现花瓣病斑像素级量化，方法新颖且可复用，对花卉病害智能监测有参考价值。",[280],{"name":272,"url":269},[28,29,30,282,283],"开源工具","植物病害检测",{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},[],{"status":41,"locked":13,"version":42,"method":43,"evidence":287},{"candidates":288,"agricultural_evidence":48},[289,293,297],{"code":34,"score":55,"supported":48,"title":290,"abstract":291,"tags":292},[],[228],[29],{"code":60,"score":20,"supported":13,"title":294,"abstract":295,"tags":296},[],[],[30],{"code":54,"score":20,"supported":13,"title":298,"abstract":299,"tags":300},[],[],[28],[302,303],"农业人工智能 植物病害检测 开源工具 智慧农业","农业人工智能 植物病害检测","农业人工智能植物病害检测开源工具智慧农业-2285","10.1016\u002Fj.compag.2026.112421",{"doi":305,"openalex_id":307,"authors":308,"venue":272,"cited_by_count":71,"oa_url":8,"card":329,"direction":76,"ingested_from":134},"W7212311236",[309,311,313,315,317,319,321,324,326],{"name":310,"orcid":8},"Yao Huang",{"name":312,"orcid":8},"Xiaoqian Cao",{"name":314,"orcid":8},"Fangqi Liu",{"name":316,"orcid":8},"Hong Sha",{"name":318,"orcid":8},"Phurisorn Watcharatpong",{"name":320,"orcid":8},"Ruoheng Jian",{"name":322,"orcid":323},"Wen Chen","https:\u002F\u002Forcid.org\u002F0000-0001-8493-0157",{"name":325,"orcid":8},"Yiqian Fu",{"name":327,"orcid":328},"Zhao Zhang","https:\u002F\u002Forcid.org\u002F0000-0002-7323-5500",{"tldr":330,"method":331,"finding":332,"direction":76,"opportunity":333},"提出开源两阶段深度学习流程PetalSpot，实现花瓣病斑像素级量化。","两阶段深度学习工作流，开源代码，用于花瓣病斑分割与量化。","该流程能对花瓣病斑进行像素级精确量化，且开源可复现。","可迁移至其他作物器官病斑量化，并融合多模态数据提升泛化性。","2026-09-13T23:30:02.086637Z",{"id":336,"title":337,"url":338,"summary":339,"summary_zh":340,"content":8,"source_name":341,"source_url":338,"published_at":342,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":343,"score_detail":344,"sources":348,"tags":350,"primary_topic":354,"secondary_topics":360,"classification":362,"search_phrases":379,"slug":382,"view_count":71,"doi":383,"paper":384,"created_at":398},5592,"Evaluation of Concept-Rank Collapse Audits for Computer Vision Agriculture Disease Classifiers Under Low-Label Regimes: An Interpretability Capacity Bound for Smallholder Fungicide Selectivity","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545193","Concept-based interpretability methods for plant-disease classifiers are typically evaluated by whether their concept activations correlate with human-annotated labels, but this criterion is silent on a prior question: whether the classifier's internal representation retains enough concept-rank structure for any audit to be informative. the present analysis formalises that prior question as an interpretability capacity bound. A classifier is modelled as a concept-rank operator whose singular values govern how much concept-level signal survives the projection from pixel space to class logits, and a collapse audit is defined as the ratio of retained concept-rank energy to total concept energy. Under a low-label regime, where concept annotations are scarce and the audit must be estimated from few labelled examples, the bound predicts that audit reliability degrades as a function of the concept-rank spectrum and the label budget. The analysis is a theoretical and methodological contribution: it derives the bound, states the conditions under which a collapse audit is identifiable, and characterises the operating characteristics of the audit as a decision procedure under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64). Power was estimated across 2000 design cells for cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 52 clusters were required for 5 percentage points at 80% power. The bound reframes concept audits as capacity tests rather than fidelity tests, with consequences for how smallholder fungicide-selectivity decisions should be supported by interpretable disease classifiers.","面向植物病害分类器的基于概念的可解释性方法，通常以其概念激活是否与人工标注标签相关作为评价标准，但该标准对一个更先前的问题保持沉默：分类器的内部表征是否保留了足够的概念秩结构，从而使任何审计具有信息量。本分析将该先前问题形式化为一个可解释性容量上界。分类器被建模为概念秩算子，其奇异值支配着从像素空间到类别logits的投影中概念级信号得以保留的程度，并将坍缩审计定义为保留的概念秩能量与总概念能量之比。在低标签情形下，概念标注稀缺且审计必须由少量标注样本估计，该上界预测审计可靠性会随概念秩谱和标签预算的函数而退化。本分析是一项理论和方法论贡献：推导出该上界，陈述坍缩审计可识别的条件，并在声明的方差结构（簇间SD = 6.0，残差SD = 4.5，组内相关系数 = 0.64）下刻画审计作为决策程序的操作特征。在2000个设计单元中，针对簇数为8、16、24和32以及效应量为3、5和8个百分点的情形估计了功效；在80%功效下达到5个百分点需要52个簇。该上界将概念审计重新框定为容量检验而非保真度检验，并对可解释病害分类器应如何支持小农杀菌剂选择性决策具有启示意义。","Zenodo (CERN European Organization for Nuclear Research)","2026-10-21T00:00:00Z",69,{"impact":225,"substance":345,"depth":16,"authority":346,"freshness":19,"relevant":20,"comment":347},21,12,"理论方法论文，提出可解释性容量上界并给出审计功效分析，对智慧农业病害识别可信度有参考价值，但属小众方法学研究、缺乏田间实证，暂不宜进入每日精选。",[349],{"name":341,"url":338},[28,29,351,352,353],"小农户","植物病害识别","可解释性",{"code":355,"slug":356,"name":357,"path":358,"parent_code":359},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[361],{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},{"status":41,"locked":13,"version":42,"method":43,"evidence":363},{"candidates":364,"agricultural_evidence":48},[365,370,375],{"code":355,"score":346,"supported":48,"title":366,"abstract":368,"tags":369},[367],"smallholder",[367],[351],{"code":34,"score":19,"supported":48,"title":371,"abstract":373,"tags":374},[372],"computer vision",[],[29,352],{"code":54,"score":20,"supported":13,"title":376,"abstract":377,"tags":378},[],[],[28],[380,381],"农业人工智能 植物病害识别 可解释性 智慧农业","农业人工智能 植物病害识别","农业人工智能植物病害识别可解释性智慧农业-5592","10.5281\u002Fzenodo.23545193",{"doi":383,"openalex_id":385,"authors":386,"venue":341,"cited_by_count":71,"oa_url":338,"card":393,"direction":76,"ingested_from":134},"W7223137323",[387,389,391],{"name":388,"orcid":8},"Akshay Ramgoolam",{"name":390,"orcid":8},"Priya Bundhoo Appadoo",{"name":392,"orcid":8},"Jean-Noël LEspérance",{"tldr":394,"method":395,"finding":396,"direction":76,"opportunity":397},"提出可解释性容量上界，评估低标签下概念秩崩溃审计的可靠性。","将分类器建模为概念秩算子，推导容量上界并做2000格功效模拟。","审计可靠性随概念秩谱与标签预算下降；5个百分点效应需52个聚类达80%功效。","可将容量上界用于小农户杀菌剂选择性决策的可解释模型审计与标签预算设计。","2026-10-22T11:21:10.425290Z",{"id":400,"title":401,"url":402,"summary":403,"summary_zh":404,"content":8,"source_name":341,"source_url":402,"published_at":342,"category":11,"cover_url":8,"hotness":19,"is_selected":13,"score":343,"score_detail":405,"sources":407,"tags":409,"primary_topic":413,"secondary_topics":419,"classification":421,"search_phrases":445,"slug":448,"view_count":71,"doi":449,"paper":450,"created_at":462},5591,"Hierarchical Conformal Risk Control for Artificial Intelligence Agriculture Irrigation Scheduling Under Unmodeled Infiltration: A Computer Safety-Guarantee Evaluation","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23543665","Irrigation scheduling algorithms that rely on soil-moisture forecasts are vulnerable to unmodeled infiltration, the process by which water bypasses the root zone through preferential pathways that resistance-based soil models cannot represent. The framework is evaluated through a replication-based design study that characterises the operating characteristics of the before-after-control-impact estimator used to audit scheduler performance, under a declared variance structure with between-cluster standard deviation 6.0, residual standard deviation 4.5 and intraclass correlation 0.64. Power was estimated across 2000 design cells spanning cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The analysis indicates that 54 clusters are required to detect a 5-percentage-point effect at 80% power, and that the minimum detectable effect declines steeply as cluster count rises from 8 to 32. The contribution is a formal safety guarantee for irrigation decision support that remains valid when the infiltration model is wrong, together with an audit design that makes the guarantee empirically checkable.","依赖土壤湿度预报的灌溉调度算法容易受到未建模入渗的影响，即水分通过优先路径绕过根区的过程，而基于阻力的土壤模型无法表征这一过程。该框架通过一项基于重复的设计研究进行评估，该研究刻画了用于审计调度器性能的前后对照影响估计量的运行特性，其所依据的声明方差结构为：组间标准差6.0、残差标准差4.5、组内相关系数0.64。功效估计覆盖2000个设计单元，涵盖聚类数8、16、24和32以及效应量3、5和8个百分点。分析表明，在80%功效下检测5个百分点效应需要54个聚类，且最小可检测效应随聚类数从8增至32而急剧下降。本文的贡献在于为灌溉决策支持提供了一种正式的安全保证，该保证在入渗模型错误时仍然有效，同时提出了一种使该保证可经实证检验的审计设计。",{"impact":225,"substance":94,"depth":16,"authority":346,"freshness":95,"relevant":20,"comment":406},"该研究提出在入渗模型失配条件下仍有效的灌溉决策安全保证框架，并给出可审计的试验设计，方法新颖、结论可靠，但属细分方法学研究，产业影响有限，适合作为专业精选而非头条。",[408],{"name":341,"url":402},[28,29,410,411,412],"灌溉决策","安全保证","试验设计",{"code":414,"slug":415,"name":416,"path":417,"parent_code":418},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation","H",[420],{"code":34,"slug":35,"name":36,"path":37,"parent_code":38},{"status":41,"locked":13,"version":42,"method":43,"evidence":422},{"candidates":423,"agricultural_evidence":48},[424,429,434,441],{"code":414,"score":215,"supported":48,"title":425,"abstract":426,"tags":428},[415],[427,415],"灌溉",[410],{"code":34,"score":346,"supported":48,"title":430,"abstract":432,"tags":433},[431],"artificial intelligence",[110],[29],{"code":435,"score":81,"supported":48,"title":436,"abstract":437,"tags":440},"H2",[],[438,439],"土壤","soil",[],{"code":54,"score":20,"supported":13,"title":442,"abstract":443,"tags":444},[],[],[28],[446,447],"灌溉调度 安全保证","农业人工智能 安全保证 智慧农业 灌溉决策","灌溉调度安全保证-5591","10.5281\u002Fzenodo.23543665",{"doi":449,"openalex_id":451,"authors":452,"venue":341,"cited_by_count":71,"oa_url":402,"card":457,"direction":76,"ingested_from":134},"W7223131205",[453,455],{"name":454,"orcid":8},"Nassima Benkhelifa",{"name":456,"orcid":8},"Yacine Aït Ahmed",{"tldr":458,"method":459,"finding":460,"direction":76,"opportunity":461},"提出分层保形风险控制，为未建模入渗下的灌溉调度提供安全保证。","分层保形风险控制与基于重复设计的BACI审计，2000个设计单元功效分析。","检测5个百分点效应需54个集群，最小可检测效应随集群数增加陡降。","可将保形风险控制扩展到其他未建模扰动下的农业决策安全保证与审计设计。","2026-10-22T11:21:10.283561Z"]