[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-13":3,"related-13":86},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":8,"published_at":11,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":23,"tags":25,"primary_topic":30,"secondary_topics":36,"classification":37,"search_phrases":80,"slug":83,"view_count":84,"doi":8,"paper":8,"created_at":85},13,"8项农业人工智能行业标准完成立项，智慧农业迈入标准化新阶段","https:\u002F\u002Fwww.toutiao.com\u002Farticle\u002F7666321638366921231\u002F","农业农村部农业信息化标准化技术委员会归口管理的8项农业人工智能行业标准完成立项，覆盖智能装备、作物智能监测、农业农村大模型、智慧农机作业、农产品溯源、低空无人机农事作业等核心细分领域，统一田间传感器数据口径、农机自动驾驶通信协议和病虫害AI识别判定标准，使不同企业产品可相互兼容、降低县域数字化改造整体成本。2026年中央一号文件将培育农业新质生产力置于突出位置，未来三年农业部门将以八项标准为核心在全国粮食主产区和特色农产品产区开展试点应用。",null,"**编者按：**2026年是“十五五”开局之年。内需市场潜力持续释放，消费新场景不断涌现，文旅融合、冰雪经济、国潮消费等热点频现，展现出超大规模市场的韧性与活力。为深入解读开局新气象，央广网推出《数读中国开局新活力》系列报道，以数据为支撑，聚焦各地消费场景、产业升级与民生改善的生动实践，呈现中国经济稳中求进、活力迸发的时代图景。\n\n![Image 1](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-axegupay5k\u002F9d562a6e7a9445b4865af57cd44518ef~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1789107844&x-signature=2K2qV1z7KUtisfMluykHCI%2FMaD8%3D)\n央广网上海7月25日消息（记者郑晓蔚 何智康）当前，人工智能引领的科技革命加速向各领域渗透，也为乡村高质量发展注入强劲动能。2026年中央一号文件明确，因地制宜发展农业新质生产力，促进人工智能与农业发展相结合。6月，由农业农村部农业信息化标准化技术委员会归口管理的8项农业人工智能相关行业标准获立项，覆盖智能装备、作物智能监测、农业农村大模型等多个领域，标志着我国农业人工智能产业迈入标准化、规范化、规模化发展新阶段。\n\n近期在上海落幕的2026世界人工智能大会暨人工智能全球治理高级别会议，集中呈现了AI赋能乡村振兴阶段化创新成果。大会主会场展示的智慧农机、农业机器人、低空植保设备等新型农事装备，已投入农业生产一线，成为提质增效的“新农具”。同时，张江科学会堂分会场举办“AI赋能乡村振兴场景”发布会，汇聚专家学者、科创企业、基层农事从业者，发布多款适配乡村发展的数字化解决方案，推动前沿科技助农项目落地生根，开花结果。\n\n**将AI应用到农业真实场景中，切实解决问题**\n\n首次参会的农业AI领域专精特新企业麦麦科技，带来全栈式农业垂类大模型落地方案。集团创始人、董事长兼CEO李楠在展会期间表示，“农业AI起步晚，但正因如此，我们有了更广阔的天地可以耕耘。当前最紧迫的任务，不是比谁的机器人跑得快、跳得高，而是把实验室里的算法和理论数据，真正应用到农业真实场景中去解决问题。”\n\n![Image 2](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-tjoges91tu\u002F4711479b5eafd7ddda08cae4c99ca414~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1789107844&x-signature=h8co4sxlzTHVGfUCxn8n6kGHQG4%3D)\n湖北荆门某植物工厂中 集成嵌入式AI算法的智能决策执行系统可对作物生长状态进行感知、研判与自适应管控（央广网发 徐登翔 摄）\n\n据了解，麦麦科技依托全国百亿级涉农数据底座、500TB多源农业数据湖，以及20亿大模型技术体系总参数，落地“Model+Robot”一体化解决方案，将定制化作物生长垂类大模型深度嵌入各类智能农机设备，让机器人摆脱“不懂农业，不懂作物”的局限，升级为可自主研判长势、管控田间生产的智能农事决策助手。\n\n湖北荆门漳河柑橘园接入该农业大模型后，告别了凭经验种植的粗放模式。AI系统可全天候监测田间环境，预判花期、汛期波动，动态智能管控农资投入，系统解决了柑橘种植病虫难控、水肥浪费、果品品质不稳定等难题。经过数字化改造，基地全年化肥农药使用量下降35%，果园亩均年增收超千元，优质果率稳定维持在85%以上，实现生态价值与经济效益“双赢”。\n\n平原农田实现智能管护，山地农耕的痛点也迎来精准破解。“我们做科技助农项目，让机器狗走进田间地头，就是希望农民伯伯们能少一些辛苦。那些不必要的劳累，能省一点是一点。”云深处科技工作人员在展会现场受访时动情表示，“我们希望机器狗可以帮农户去危险的山地，扛下爬坡、负重、运输的险活累活，让农耕更安全、更省力。”\n\n![Image 3](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-tjoges91tu\u002F251f7b65af9fc74955733dc4cae58ffa~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1789107844&x-signature=TGDEiyalUCja4p6aGN4%2FFYfGl5o%3D)\n机器狗可去往山地田间完成鲜果运输（央广网发 李婷 摄）\n\n云深处科技推出的轮足机器人可平稳穿行45度碎石泥泞山路，完成鲜果转运，适配仙居杨梅、四川茶园、福建果园等山地作业场景，实实在在为山区农户减负增效。\n\n“本次展会让我们看到了乡村科创的无限可能，很多企业其实已深耕农业细分领域多年、潜心攻克技术难题。”一位涉足农业赛道多年的参展企业高管直呼“来值了”，“各类农业科创技术在这里交流融汇，未来一定会全方位赋能农业发展。”\n\n对于智慧农业的普及趋势，北航交通学院汽车工程系主任王建作出预判：“播种、植保、巡田、采收这些重复性高强度农活，未来都会被各类机器人全面承接。3至5年内，远程AI管控、低空无人机规模化巡田就能在全国多数县域落地普及，新农人只需远程操控设备就能完成田间管护。”\n\n**开发非遗文创，推动乡土文化保护传承与活化利用**\n\n智能种养筑牢乡村产业基本盘，乡土非遗文化也在AI加持下跳出展馆库房，打造出乡村特色差异化发展新亮点。\n\n长期以来，昭通甲马、纳西东巴纸、海派绒绣等诸多乡村非遗，深陷“传承难、变现难”的困境。传承人老龄化、市场渠道窄，大量传统技艺只能静态陈列，难以转化为村民增收的实体产业。本次“AI赋能乡村振兴场景”发布会上，上海工程技术大学校长、瑞典皇家工程科学院院士娄永琪牵头联合TEMOINS文化智库平台，为 AI非遗融合创新实验室揭牌。实验室依托校企协同创新机制，运用AI技术开发非遗文创产品，以数字化手段推动乡土文化保护传承与活化利用，同步拓宽村民增收渠道。\n\n![Image 4](https:\u002F\u002Fp3-sign.toutiaoimg.com\u002Ftos-cn-i-tjoges91tu\u002F244828a0fdf1ff1022ca9fa657fc01f8~tplv-tt-origin-web:gif.jpeg?_iz=58558&from=article.pc_detail&lk3s=953192f4&x-expires=1789107844&x-signature=3DUzQp9qu1Kl7ztqXvf5GVEbsL8%3D)\n在2026世界人工智能大会期间 融入AI元素的昭通甲马吸引年轻人目光（央广网记者 何智康 摄）\n\n**打通流通堵点，让每一份乡村好物被看见、被选择**\n\n长期制约乡村产业提质增收的核心症结，还存在于传统流通渠道层级冗余、市场预判滞后、消费信任不足，本次“AI赋能乡村振兴场景”发布会上，多家产业平台负责人聚焦农产品产销堵点，共探AI破局之道。\n\n梦饷科技联合创始人兼CEO冷静现场用一组数据展示了AI赋能农产品大促的威力：在山东蒙阴，利用AI提前72小时预测爆款需求，实施“按单采收”，帮助千年桃乡两小时卖出20万颗黄桃；在黑龙江五常，通过全域数据分析预判“高潜”爆品，配合店主原产地溯源，一周内售出260吨大米，相当于一家超市5年的销量……依托自研智能供应链操作系统，梦饷科技打造AI赋能农产品产销的成熟落地模式，打破传统农业“盲种、盲收、盲卖”的粗放发展路径。\n\n冷静表示，AI已成为赋能乡村产业的核心生产力，团队将持续深耕智能技术应用，以AI之力点亮乡村振兴的每一个角落，让每一份乡村好物被看见、被信任、被选择。\n\n“2026年中央一号文件释出的政策信号已经非常清晰——农业AI的规模化落地，正从‘可选项’变为‘必选项’。我们将潜心布局，借助AI技术让农业生产更高效、让乡村发展更有活力。”一位农业领域参展企业负责人受访时表示。\n\n“十五五”开局之年，农业AI已扎根田间地头、融入乡村肌理，成为惠及三农、造福民生的“硬核刚需”。随着农业AI行业标准持续完善、涉农数据底座不断夯实、智能技术不断迭代，数字新质生产力将持续下沉县域乡村，全方位赋能农业生产、文化传承、产销流通全链条。一幅农耕高效、产业兴旺、宜居宜业、共同富裕的和美乡村新“丰”景，正徐徐展开。\n\n**更多精彩资讯请在应用市场下载“央广网”客户端。欢迎提供新闻线索，24小时报料热线400-800-0088；消费者也可通过央广网“啄木鸟消费者投诉平台”线上投诉。版权声明：本文章版权归属央广网所有，未经授权不得转载。转载请联系：cnrbanquan@cnr.cn，不尊重原创的行为我们将追究责任。**","央广网 \"数读中国开局新活力\"系列 · 2026-07-25","2026-07-25T00:00:00Z","政策",10,true,82,{"impact":17,"substance":18,"depth":19,"authority":20,"freshness":21,"relevant":21,"comment":22},28,22,17,14,1,"全国性农业AI标准立项，标志产业进入标准化新阶段，内容详实，但发布已超一月，时效性低。",[24],{"name":10,"url":6},[26,27,28,29],"智慧农业","农业人工智能","乡村振兴","行业标准",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":38,"locked":39,"version":40,"method":41,"evidence":42},"classified",false,"2026-10-22.2","rules",{"candidates":43,"agricultural_evidence":14},[44,50,57,64,70,75],{"code":31,"score":45,"supported":14,"title":46,"abstract":47,"tags":49},15,[27],[27,48],"大模型",[27],{"code":51,"score":52,"supported":14,"title":53,"abstract":54,"tags":56},"I1",12,[26],[55],"农业信息化",[26],{"code":58,"score":59,"supported":14,"title":60,"abstract":61,"tags":63},"E4",3,[],[62],"农产品溯源",[],{"code":65,"score":59,"supported":14,"title":66,"abstract":67,"tags":69},"F5",[],[68],"农机",[],{"code":71,"score":21,"supported":39,"title":72,"abstract":73,"tags":74},"G",[],[],[28],{"code":76,"score":21,"supported":39,"title":77,"abstract":78,"tags":79},"J1",[],[],[29],[81,82],"农业人工智能 乡村振兴 智慧农业 行业标准","农业人工智能 乡村振兴","农业人工智能乡村振兴智慧农业行业标准-13",16,"2026-07-27T05:05:33.370140Z",{"total":87,"page":21,"page_size":87,"items":88},6,[89,161,209,269,335,387],{"id":90,"title":91,"url":92,"summary":93,"summary_zh":94,"content":8,"source_name":95,"source_url":92,"published_at":96,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":98,"score_detail":99,"sources":104,"tags":106,"primary_topic":110,"secondary_topics":111,"classification":112,"search_phrases":139,"slug":142,"view_count":143,"doi":144,"paper":145,"created_at":160},2650,"Automated Cocoa Bean Classification and Defect Detection Based on ASEAN Standard Using Image Processing and Feature-Based Analysis","https:\u002F\u002Fdoi.org\u002F10.64823\u002Fijeee.2601005","This study presents an automated cocoa bean classification and defect detection system developed in accordance with the ASEAN Standard for Cocoa Bean (ASEAN Stan 34:2014). Implemented in MATLAB, the system utilizes digital image processing and computer vision techniques to segment individual cocoa beans, extract geometric, color, and texture features, and classify them into quality categories: Extra Class, Class I, Class II, and Non-Compliant. The image processing workflow integrates color space transformations (RGB to HSV), adaptive thresholding, and morphological filtering to ensure accurate segmentation. Feature extraction targets parameters such as area, length, aspect ratio, texture entropy, and color uniformity to identify specific defects including moldy, slaty, insect-damaged, and germinated beans. Validation against independent expert grading achieved an overall reliability rate of 96.2%, with high precision (0.93) and recall (0.91), and an average processing speed of 1.3 seconds per image. The developed analyzer provides an objective, rapid, and repeatable tool to support standardization and postharvest cocoa quality control across the ASEAN region. Keywords: Defect Detection; MATLAB; cocoa bean grading; ASEAN Standard; digital image processing; computer vision","本研究提出了一套依据《东盟可可豆标准》（ASEAN Stan 34:2014）开发的自动化可可豆分类与缺陷检测系统。该系统在MATLAB中实现，利用数字图像处理与计算机视觉技术对单颗可可豆进行分割，提取几何、颜色和纹理特征，并将其分为优质级（Extra Class）、一级（Class I）、二级（Class II）和不合格（Non-Compliant）四个质量类别。图像处理流程整合了颜色空间变换（RGB转HSV）、自适应阈值分割和形态学滤波，以确保分割的准确性。特征提取针对面积、长度、长宽比、纹理熵和颜色均匀性等参数，以识别霉变、僵化、虫害和发芽等特定缺陷。经与独立专家评级进行验证，系统总体可靠率达到96.2%，精确率为0.93，召回率为0.91，平均处理速度为每张图像1.3秒。所开发的分析仪为支持东盟地区可可采后质量控制的标准化提供了一种客观、快速且可重复的工具。关键词：缺陷检测；MATLAB；可可豆分级；东盟标准；数字图像处理；计算机视觉","International Journal of Electrical and Electronics Engineering","2026-09-14T00:00:00Z","论文",77,{"impact":100,"substance":101,"depth":19,"authority":4,"freshness":102,"relevant":21,"comment":103},18,21,8,"基于东盟标准的可可豆自动分级与缺陷检测研究，方法成熟、验证可靠（96.2%），对热带经济作物产后质控有实用价值，但属细分领域技术进展，影响力有限。",[105],{"name":95,"url":92},[26,27,107,29,108,109],"计算机视觉","农产品分级","可可产业",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[],{"status":38,"locked":39,"version":40,"method":41,"evidence":113},{"candidates":114,"agricultural_evidence":14},[115,120,126,131,135],{"code":31,"score":102,"supported":14,"title":116,"abstract":117,"tags":119},[],[107,118],"computer vision",[27,107],{"code":121,"score":59,"supported":14,"title":122,"abstract":123,"tags":125},"B3",[],[124],"可可",[],{"code":127,"score":21,"supported":39,"title":128,"abstract":129,"tags":130},"E1",[],[],[108],{"code":51,"score":21,"supported":39,"title":132,"abstract":133,"tags":134},[],[],[26],{"code":76,"score":21,"supported":39,"title":136,"abstract":137,"tags":138},[],[],[29],[140,141],"农业人工智能 农产品分级 计算机视觉 可可产业","农业人工智能 农产品分级","农业人工智能农产品分级计算机视觉可可产业-2650",2,"10.64823\u002Fijeee.2601005",{"doi":144,"openalex_id":146,"authors":147,"venue":95,"cited_by_count":150,"oa_url":151,"card":152,"direction":158,"ingested_from":159},"W7212670192",[148],{"name":149,"orcid":8},"Jennifer Natnat",0,"https:\u002F\u002Fioro.org\u002Farticle\u002F529160740839\u002Fpdf",{"tldr":153,"method":154,"finding":155,"direction":156,"opportunity":157},"基于图像处理与特征分析，按东盟标准自动分类可可豆并检测缺陷。","MATLAB图像处理，RGB转HSV、自适应阈值、形态学滤波，提取几何颜色纹理特","系统分类准确率96.2%，精度0.93，召回0.91，单图处理1.3秒。","农业人工智能与决策模型","可扩展至多作物、多标准实时分级，结合深度学习提升复杂缺陷识别鲁棒性。","智慧农业 \u002F 农业物联网","openalex","2026-09-16T23:30:13.899726Z",{"id":162,"title":163,"url":164,"summary":165,"summary_zh":8,"content":8,"source_name":166,"source_url":8,"published_at":167,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":168,"score_detail":169,"sources":173,"tags":175,"primary_topic":178,"secondary_topics":179,"classification":180,"search_phrases":196,"slug":199,"view_count":21,"doi":8,"paper":200,"created_at":208},1588,"Expert Panel | The Rapid Rise of AI in Agriculture: A Multi-Stakeholder Discussion","https:\u002F\u002Fwww.smartagrichain.com\u002Fnews\u002F1003\u002Fexpert-panel-the-rapid-rise-of-ai-in-agriculture-a-multi-stakeholder-discussion","MDPI《Smart Agricultural Technology》发表专家圆桌：人工智能正以多种形式渗入农业全产业链。本期邀请来自荷兰瓦赫宁根大学、中国农业大学、美国农业部农业研究服务局与以色列农业研究组织的多位专家，围绕大模型与垂直模型在农业中的落地路径展开多利益相关方讨论，重点议题包括基础模型与生成式 AI 决策支持、农业垂直大模型在病虫害识别与产量预测上的差异，及其在不同地域、不同规模经营主体中的适用性。讨论结论认为，建立覆盖模型可解释性、田间验证、跨域迁移和共享开放数据的统一评测标准将是下一阶段推进农业 AI 落地的关键。","MDPI Smart Agricultural Technology","2026-09-02T00:00:00Z",81,{"impact":18,"substance":170,"depth":100,"authority":20,"freshness":171,"relevant":21,"comment":172},20,7,"专家圆桌讨论农业AI落地关键议题，提出统一评测标准，具有行业前瞻性。",[174],{"name":166,"url":164},[26,27,29,176,177],"基础模型","多利益相关方",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[],{"status":38,"locked":39,"version":40,"method":41,"evidence":181},{"candidates":182,"agricultural_evidence":14},[183,188,192],{"code":31,"score":171,"supported":14,"title":184,"abstract":185,"tags":187},[],[48,186],"病虫害识别",[27],{"code":51,"score":21,"supported":39,"title":189,"abstract":190,"tags":191},[],[],[26],{"code":76,"score":21,"supported":39,"title":193,"abstract":194,"tags":195},[],[],[29],[197,198],"农业人工智能 多利益相关方 基础模型 智慧农业","农业人工智能 多利益相关方","农业人工智能多利益相关方基础模型智慧农业-1588",{"doi":8,"openalex_id":8,"authors":201,"venue":8,"cited_by_count":150,"oa_url":8,"card":202,"direction":156,"ingested_from":207},[],{"tldr":203,"method":204,"finding":205,"direction":156,"opportunity":206},"专家圆桌讨论AI在农业中的应用，强调建立统一评测标准以推进落地。","多利益相关方专家讨论，聚焦大模型与垂直模型在农业的应用。","统一评测标准是农业AI落地的关键，涵盖可解释性、验证、迁移和数据共享。","可研究农业AI模型的跨域迁移与统一评测标准，特别是小规模经营主体的适用性。","agent","2026-09-04T00:05:34.570172Z",{"id":210,"title":211,"url":212,"summary":213,"summary_zh":214,"content":8,"source_name":215,"source_url":212,"published_at":216,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":217,"score_detail":218,"sources":220,"tags":222,"primary_topic":226,"secondary_topics":231,"classification":233,"search_phrases":249,"slug":252,"view_count":150,"doi":253,"paper":254,"created_at":268},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":102,"substance":101,"depth":100,"authority":52,"freshness":13,"relevant":21,"comment":219},"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。",[221],{"name":215,"url":212},[26,27,223,224,225],"小农户","植物病害识别","可解释性",{"code":227,"slug":228,"name":229,"path":230,"parent_code":71},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services",[232],{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},{"status":38,"locked":39,"version":40,"method":41,"evidence":234},{"candidates":235,"agricultural_evidence":14},[236,241,245],{"code":227,"score":52,"supported":14,"title":237,"abstract":239,"tags":240},[238],"smallholder",[238],[223],{"code":31,"score":13,"supported":14,"title":242,"abstract":243,"tags":244},[118],[],[27,224],{"code":51,"score":21,"supported":39,"title":246,"abstract":247,"tags":248},[],[],[26],[250,251],"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":253,"openalex_id":255,"authors":256,"venue":215,"cited_by_count":150,"oa_url":212,"card":263,"direction":156,"ingested_from":159},"W7223137323",[257,259,261],{"name":258,"orcid":8},"Akshay Ramgoolam",{"name":260,"orcid":8},"Priya Bundhoo Appadoo",{"name":262,"orcid":8},"Jean-Noël LEspérance",{"tldr":264,"method":265,"finding":266,"direction":156,"opportunity":267},"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":270,"title":271,"url":272,"summary":273,"summary_zh":274,"content":8,"source_name":215,"source_url":272,"published_at":216,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":217,"score_detail":275,"sources":278,"tags":280,"primary_topic":284,"secondary_topics":290,"classification":292,"search_phrases":317,"slug":320,"view_count":150,"doi":321,"paper":322,"created_at":334},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":102,"substance":18,"depth":100,"authority":52,"freshness":276,"relevant":21,"comment":277},9,"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。",[279],{"name":215,"url":272},[26,27,281,282,283],"灌溉决策","安全保证","试验设计",{"code":285,"slug":286,"name":287,"path":288,"parent_code":289},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation","H",[291],{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},{"status":38,"locked":39,"version":40,"method":41,"evidence":293},{"candidates":294,"agricultural_evidence":14},[295,300,306,313],{"code":285,"score":45,"supported":14,"title":296,"abstract":297,"tags":299},[286],[298,286],"灌溉",[281],{"code":31,"score":52,"supported":14,"title":301,"abstract":303,"tags":305},[302],"artificial intelligence",[304],"decision support",[27],{"code":307,"score":87,"supported":14,"title":308,"abstract":309,"tags":312},"H2",[],[310,311],"土壤","soil",[],{"code":51,"score":21,"supported":39,"title":314,"abstract":315,"tags":316},[],[],[26],[318,319],"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":321,"openalex_id":323,"authors":324,"venue":215,"cited_by_count":150,"oa_url":272,"card":329,"direction":156,"ingested_from":159},"W7223131205",[325,327],{"name":326,"orcid":8},"Nassima Benkhelifa",{"name":328,"orcid":8},"Yacine Aït Ahmed",{"tldr":330,"method":331,"finding":332,"direction":156,"opportunity":333},"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",{"id":336,"title":337,"url":338,"summary":339,"summary_zh":340,"content":8,"source_name":215,"source_url":338,"published_at":216,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":341,"score_detail":342,"sources":344,"tags":346,"primary_topic":349,"secondary_topics":350,"classification":352,"search_phrases":367,"slug":370,"view_count":150,"doi":371,"paper":372,"created_at":386},5590,"Evaluation of Concept-Probe Temporal-Stability Audits for Computer Vision Agriculture Disease Classifiers Under Seasonal Phenology Drift: An Interpretability Decay Bound for Smallholder Fungicide Timing","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545234","Concept-probe audits are increasingly used to certify that a crop-disease classifier's internal representations track the pathological concepts a smallholder would act on, but a probe certified at one point in the season may silently decouple from those concepts as phenology shifts the visual statistics of the field. the present analysis treats that decoupling as a formal object. It defines a temporal-stability audit as a sequence of concept-probe accuracies measured across defined phenological windows, and it derives an interpretability decay bound: the rate at which probe accuracy can fall before the fungicide-timing decision it is meant to support becomes unreliable. The bound is expressed as a function of the probe's initial margin, the drift in concept prevalence between windows, and the classifier's sensitivity to the drifted features, and it yields a certified audit interval rather than a single-point certificate. The paper then characterises the operating characteristics of the audit design under a declared variance structure, using a replication-based evaluation of the before-after-control-impact estimator with between-cluster SD = 6.0, residual SD = 4.5 and intraclass correlation = 0.64. Power was estimated across 2000 design cells over cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 54 clusters were required for 5 percentage points at 80% power. The contribution is a decision-theoretic account of when a concept-probe certificate expires, and a design rule for scheduling re-audits across the season.","概念探针审计（concept-probe audit）日益被用于验证作物病害分类器的内部表征是否追踪小农户会据以采取行动的病理学概念，但在生长季某一时点获得认证的探针，可能随着物候变化改变田间的视觉统计特征而悄然与这些概念脱钩。本分析将这种脱钩视为一个形式化对象。它将时间稳定性审计（temporal-stability audit）定义为在既定物候窗口内测量的一系列概念探针准确率，并推导出一个可解释性衰减界限（interpretability decay bound）：即探针准确率在其所支持的杀菌剂施用时机决策变得不可靠之前所能下降的速率。该界限表示为探针初始裕度、窗口间概念流行率漂移以及分类器对漂移特征敏感度的函数，并给出一个经认证的审计区间而非单点认证。随后，本文在声明的方差结构下刻画该审计设计的运行特征，采用基于重复的前后对照影响估计量评估，其中组间标准差 = 6.0、残差标准差 = 4.5、组内相关系数 = 0.64。功效在 2000 个设计单元上估计，聚类数取 8、16、24 和 32，效应量取 3、5 和 8 个百分点；在 80% 功效下达到 5 个百分点需要 54 个聚类。本文的贡献在于从决策论角度阐明概念探针认证何时失效，并给出在整个生长季内安排重新审计的设计规则。",70,{"impact":102,"substance":18,"depth":100,"authority":52,"freshness":13,"relevant":21,"comment":343},"该论文提出概念探针审计的时序稳定性框架与可解释性衰减界，方法新颖、统计设计扎实，但属小众方法论研究，产业影响有限，适合作为专业精选而非头条。",[345],{"name":215,"url":338},[26,27,223,347,348],"作物病害识别","模型审计",{"code":227,"slug":228,"name":229,"path":230,"parent_code":71},[351],{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},{"status":38,"locked":39,"version":40,"method":41,"evidence":353},{"candidates":354,"agricultural_evidence":14},[355,359,363],{"code":227,"score":45,"supported":14,"title":356,"abstract":357,"tags":358},[238],[223,238],[223],{"code":31,"score":13,"supported":14,"title":360,"abstract":361,"tags":362},[118],[],[347,27],{"code":51,"score":21,"supported":39,"title":364,"abstract":365,"tags":366},[],[],[26],[368,369],"概念探针 时序稳定性 作物病害分类器","杀菌剂 施药时机 可解释性衰减","概念探针时序稳定性作物病害分类器-5590","10.5281\u002Fzenodo.23545234",{"doi":371,"openalex_id":373,"authors":374,"venue":215,"cited_by_count":150,"oa_url":338,"card":381,"direction":156,"ingested_from":159},"W7223088030",[375,377,379],{"name":376,"orcid":8},"Abdi-Basid Idriss Farah",{"name":378,"orcid":8},"Amina Abdillahi Guedi",{"name":380,"orcid":8},"Moussa Omar Youssouf",{"tldr":382,"method":383,"finding":384,"direction":156,"opportunity":385},"提出概念探针时间稳定性审计与可解释性衰减界，判断何时需重新审计。","定义跨物候窗口探针精度序列，推导衰减界，用BACI复制评估统计功效。","5个百分点效应需54个聚类达80%功效，审计证书会随物候漂移过期。","可将衰减界与再审计调度规则嵌入小农户杀菌剂时机决策支持系统，并验证跨作物泛化。","2026-10-22T11:21:08.978348Z",{"id":388,"title":389,"url":390,"summary":391,"summary_zh":392,"content":8,"source_name":215,"source_url":390,"published_at":216,"category":97,"cover_url":8,"hotness":13,"is_selected":39,"score":393,"score_detail":394,"sources":396,"tags":398,"primary_topic":401,"secondary_topics":402,"classification":403,"search_phrases":421,"slug":424,"view_count":150,"doi":425,"paper":426,"created_at":436},5589,"Conformal-Risk Evaluation for Artificial Intelligence Agriculture Spray-Nozzle Fault Detection Under Label-Delay Feedback: A Computer Coverage Audit","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23544107","Spray-nozzle fault detection in precision agriculture is a classification problem in which the label that confirms a fault arrives after the spray pass, so the detector must act on unlabelled evidence and its error rate must be certified before the label exists. Conformal prediction offers distribution-free coverage guarantees, but the guarantee is stated for exchangeable calibration and test data, an assumption that label-delayed feedback violates because the calibration set is always older than the deployment set. the present analysis formalises label-delay feedback as a distribution shift in the conformity score and derives a coverage bound for split-conformal fault detectors under that shift. The bound separates the nominal miscoverage rate into an exchangeable component and a delay-induced component that scales with the total variation distance between the calibration and deployment score distributions. A replication-based evaluation of the ['Intercept', 'exposure', 'group', 'severity'] estimator was conducted: 600 data sets of N = 500 were drawn from a declared data-generating process (intercept 12.0, exposure 0.42, group -3.1, severity 1.35, error SD 4.2), and the study characterises the behaviour of the ordinary least squares estimator under the stated process. The estimator recovered the declared coefficients with estimated bias 0.01122 for the intercept and -0.0002 for exposure, root mean squared error 0.9217 and 0.01614 respectively, and empirical coverage of the nominal 95% intervals of 1.0 for both parameters across 600 replications. The coverage audit shows that the delay-induced term, not the exchangeable term, dominates the risk of a conformal spray-nozzle detector, and that the audit procedure can certify a detector before field labels return.","在精准农业中，喷雾喷嘴故障检测是一个分类问题，其中确认故障的标签在喷雾作业结束后才到达，因此检测器必须基于未标记的证据采取行动，并且其错误率必须在标签存在之前得到认证。保形预测（conformal prediction）提供了无分布覆盖保证，但该保证是针对可交换的校准数据和测试数据而言的，而标签延迟反馈违反了这一假设，因为校准集总是比部署集更旧。本分析将标签延迟反馈形式化为一致性分数中的分布偏移，并推导了在该偏移下分裂保形故障检测器的覆盖界。该界将名义错误覆盖 rate 分解为可交换分量和延迟诱导分量，后者随校准与部署分数分布之间的全变差距离而缩放。对 ['Intercept', 'exposure', 'group', 'severity'] 估计器进行了基于重复的评估：从声明的数据生成过程（截距 12.0，暴露 0.42，分组 -3.1，严重程度 1.35，误差标准差 4.2）中抽取了 600 个 N = 500 的数据集，研究刻画了普通最小二乘估计器在所述过程下的行为。该估计器恢复了声明的系数，截距的估计偏差为 0.01122，暴露为 -0.0002，均方根误差分别为 0.9217 和 0.01614，两个参数在 600 次重复中名义 95% 区间的经验覆盖率为 1.0。覆盖审计表明，延迟诱导项而非可交换项主导了保形喷雾喷嘴检测器的风险，并且该审计程序可以在田间标签返回之前对检测器进行认证。",62,{"impact":102,"substance":100,"depth":19,"authority":276,"freshness":13,"relevant":21,"comment":395},"论文提出标签延迟下共形预测覆盖界，方法有新意但属预印本、应用面窄，可作为专业补充而非每日精选头条。",[397],{"name":215,"url":390},[26,27,399,400],"精准施药","病虫害监测",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[],{"status":38,"locked":39,"version":40,"method":41,"evidence":404},{"candidates":405,"agricultural_evidence":14},[406,410,416],{"code":31,"score":276,"supported":14,"title":407,"abstract":408,"tags":409},[302],[],[27],{"code":51,"score":171,"supported":14,"title":411,"abstract":412,"tags":415},[],[413,414],"精准农业","precision agriculture",[26],{"code":417,"score":21,"supported":39,"title":418,"abstract":419,"tags":420},"F4",[],[],[399],[422,423],"农业人工智能 病虫害监测 智慧农业 精准施药","农业人工智能 病虫害监测","农业人工智能病虫害监测智慧农业精准施药-5589","10.5281\u002Fzenodo.23544107",{"doi":425,"openalex_id":427,"authors":428,"venue":215,"cited_by_count":150,"oa_url":390,"card":431,"direction":156,"ingested_from":159},"W7223116087",[429],{"name":430,"orcid":8},"Achieng' Atim Okello",{"tldr":432,"method":433,"finding":434,"direction":156,"opportunity":435},"将标签延迟反馈建模为共形分数分布偏移，推导覆盖率界并审计喷雾喷嘴故障检测器。","分裂共形预测、总变差距离、600次重复模拟（N=500）评估OLS估计。","延迟诱导项而非可交换项主导检测器风险，审计可在田间标签返回前认证检测器。","可将该覆盖率审计框架扩展到其他标签延迟的农业传感器故障检测，并研究延迟分布估计误差对认证的影响。","2026-10-22T11:21:02.845718Z"]