[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-3159":3,"related-3159":106},{"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,"primary_topic":9,"secondary_topics":31,"classification":32,"search_phrases":75,"slug":78,"view_count":79,"doi":80,"paper":81,"created_at":105},3159,"A multi-agent multi-modal LLM with Monte Carlo tree search for interpretable plant disease classification","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1937679","In plant disease classification tasks, existing task specific models for plant disease classification typically achieve high accuracy by employing tailored network architectures. However, these models are unable to provide reasoning chains, which reduces interpretability and undermines trustworthiness in practical applications. In addition, although multimodal large language models (MLLMs) have made significant progress in smart agriculture, their disease classification accuracy lags behind that of task-specific models due to insufficient domain knowledge. While fine-tuning can enhance domain-specific knowledge, relying solely on unimodal data or simple multimodal information remains insufficient for adequately improving classification accuracy. Specifically, the simple multimodal information in this context refers to a single global description of the entire image produced by an MLLM. To address these issues, this paper proposes a multi-agent collaborative classification framework that integrates Monte Carlo Tree Search (MCTS) with multimodal enhancement. With an MLLM as the core, the framework first introduces an MCTSguided multi-agent collaboration mechanism that directs the agents to progressively generate region-specific descriptive textual information, which together with the image forms complex multimodal information and thereby boosts classification performance. Meanwhile, the search trajectories generated by MCTS are leveraged to construct reasoning processes for enhanced interpretability. Finally, a self-training mechanism is adopted to iteratively optimize the model, further improving overall performance. Experimental results show that the framework improves the baseline MLLM’s accuracy while providing interpretable reasoning chains, achieving gains from 0.6701 to 0.9965 on 3-class potato, 0.5340 to 0.9651 on 5-class potato, 0.5387 to 0.9467 on tomato, and 0.4705 to 0.9948 on grape.","在植物病害分类任务中，现有的植物病害分类专用模型通常通过采用定制化的网络架构来实现高准确率。然而，这些模型无法提供推理链，这降低了可解释性，并在实际应用中削弱了可信度。此外，尽管多模态大语言模型（MLLMs）在智慧农业领域取得了显著进展，但由于领域知识不足，其病害分类准确率仍落后于专用模型。虽然微调可以增强领域特定知识，但仅依赖单模态数据或简单的多模态信息仍不足以充分提升分类准确率。具体而言，此处的简单多模态信息是指MLLM对整个图像生成的单一全局描述。为解决上述问题，本文提出了一种融合蒙特卡洛树搜索（MCTS）与多模态增强的多智能体协同分类框架。该框架以MLLM为核心，首先引入MCTS引导的多智能体协同机制，引导各智能体逐步生成区域特定的描述性文本信息，这些信息与图像共同构成复杂的多模态信息，从而提升分类性能。同时，利用MCTS生成的搜索轨迹构建推理过程，以增强可解释性。最后，采用自训练机制迭代优化模型，进一步提升整体性能。实验结果表明，该框架在提升基线MLLM准确率的同时提供了可解释的推理链，在三分类马铃薯上从0.6701提升至0.9965，在五分类马铃薯上从0.5340提升至0.9651，在番茄上从0.5387提升至0.9467，在葡萄上从0.4705提升至0.9948。",null,"Frontiers in Plant Science","2026-09-21T00:00:00Z","论文",10,false,80,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,14,8,1,"提出多智能体+MCTS+多模态增强的可解释植物病害分类框架，准确率大幅提升，方法新颖且结论可靠，对智慧农业AI应用有参考价值。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","农业人工智能","可解释AI","植物病害识别","多模态大模型",[],{"status":33,"locked":14,"version":34,"method":35,"evidence":36},"needs_review","2026-10-22.2","rules",{"candidates":37,"agricultural_evidence":41,"reason":74},[38,46,54,61,68],{"code":39,"score":40,"supported":41,"title":42,"abstract":43,"tags":45},"I1",7,true,[],[26,44],"smart agriculture",[26],{"code":47,"score":48,"supported":41,"title":49,"abstract":50,"tags":53},"A5",6,[],[51,52],"马铃薯","potato",[],{"code":55,"score":48,"supported":41,"title":56,"abstract":57,"tags":60},"B1",[],[58,59],"番茄","tomato",[],{"code":62,"score":48,"supported":41,"title":63,"abstract":64,"tags":67},"B2",[],[65,66],"葡萄","grape",[],{"code":69,"score":70,"supported":14,"title":71,"abstract":72,"tags":73},"I2",3,[],[],[27,28,30,29],"多个产业证据相当，需要确认主归属",[76,77],"多智能体 蒙特卡洛树搜索 植物病害分类","多模态大模型 马铃薯 番茄 葡萄 病害识别","多智能体蒙特卡洛树搜索植物病害分类-3159",2,"10.3389\u002Ffpls.2026.1937679",{"doi":80,"openalex_id":82,"authors":83,"venue":10,"cited_by_count":96,"oa_url":6,"card":97,"direction":103,"ingested_from":104},"W7213924931",[84,87,89,91,94],{"name":85,"orcid":86},"Ridong Wu","https:\u002F\u002Forcid.org\u002F0000-0003-0729-3099",{"name":88,"orcid":9},"Hua Zou",{"name":90,"orcid":9},"Zhijie Li",{"name":92,"orcid":93},"Qian Zhou","https:\u002F\u002Forcid.org\u002F0000-0001-7964-8130",{"name":95,"orcid":9},"一子 江﨑",0,{"tldr":98,"method":99,"finding":100,"direction":101,"opportunity":102},"提出多智能体多模态LLM结合蒙特卡洛树搜索，实现可解释的植物病害分类。","MCTS引导多智能体生成区域描述文本，构建复杂多模态信息并自训练优化。","分类准确率大幅提升，如马铃薯三分类从0.6701升至0.9965，并提供推理链。","农业人工智能与决策模型","可探索将MCTS多智能体框架迁移至其他作物病害或田间复杂场景，提升可解释性与泛化性。","智慧农业 \u002F 农业物联网","openalex","2026-09-22T23:30:11.045394Z",{"total":48,"page":21,"page_size":48,"items":107},[108,171,230,276,324,385],{"id":109,"title":110,"url":111,"summary":112,"summary_zh":113,"content":9,"source_name":114,"source_url":111,"published_at":115,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":116,"score_detail":117,"sources":122,"tags":124,"primary_topic":126,"secondary_topics":131,"classification":132,"search_phrases":154,"slug":157,"view_count":70,"doi":158,"paper":159,"created_at":170},4907,"PEAS-DeiT: Multi-Scale Spatial Attention Enhanced DeiT for Pea Leaf Disease Classification","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fapp162010113","Early and accurate detection of pea diseases is essential for reducing crop losses and supporting precision agriculture. Although deep learning methods have shown good performance in plant disease classification, CNN-based models may have limitations in learning long-range relationships and detailed disease features. In this study, an explainable transformer-based framework, named PEAS-DeiT, is proposed for automated pea leaf disease classification. The proposed approach integrates a multi-scale spatial attention mechanism into the Data-efficient Image Transformer (DeiT) architecture to enhance the representation of disease-related visual features. A unified pea leaf disease dataset (UPLDD) was constructed by combining multiple publicly available datasets, covering six classes, including five disease categories and healthy samples. The proposed model was evaluated against CNN-based and transformer-based architectures using five-fold cross-validation. Furthermore, explainable artificial intelligence (XAI) analysis was performed to investigate the regions contributing to model predictions. Experimental results demonstrate that PEAS-DeiT achieved an accuracy of 93.11%, showing competitive performance compared with the evaluated CNN and transformer-based models. The gradient-based saliency analysis also revealed localized response patterns in several evaluated samples, providing qualitative insight into the image regions contributing to model predictions. Statistical analyses were conducted to evaluate the performance differences among the compared models. These findings indicate that PEAS-DeiT provides a competitive approach for pea leaf disease classification within the evaluated experimental setting.","豌豆病害的早期准确检测对于减少作物损失和支持精准农业至关重要。尽管深度学习方法在植物病害分类中表现出良好性能，但基于卷积神经网络（CNN）的模型在学习长程关系和病害细节特征方面可能存在局限。本研究提出了一种基于Transformer的可解释框架PEAS-DeiT，用于豌豆叶片病害的自动化分类。该方法将多尺度空间注意力机制集成到数据高效图像Transformer（DeiT）架构中，以增强病害相关视觉特征的表征能力。通过整合多个公开数据集，构建了统一的豌豆叶片病害数据集（UPLDD），涵盖六个类别，包括五类病害和健康样本。采用五折交叉验证，将所提模型与基于CNN和基于Transformer的架构进行了评估比较。此外，进行了可解释人工智能（XAI）分析，以探究对模型预测有贡献的区域。实验结果表明，PEAS-DeiT达到了93.11%的准确率，与所评估的基于CNN和基于Transformer的模型相比展现出有竞争力的性能。基于梯度的显著性分析还揭示了若干评估样本中的局部响应模式，为理解对模型预测有贡献的图像区域提供了定性见解。通过统计分析评估了所比较模型之间的性能差异。这些发现表明，在所评估的实验设置下，PEAS-DeiT为豌豆叶片病害分类提供了一种有竞争力的方法。","Applied Sciences","2026-10-13T00:00:00Z",64,{"impact":20,"substance":17,"depth":118,"authority":119,"freshness":120,"relevant":21,"comment":121},16,13,9,"提出多尺度空间注意力增强的DeiT模型用于豌豆叶病害分类，构建六类统一数据集并达93.11%准确率，方法有创新但属实验室验证阶段，产业影响有限。",[123],{"name":114,"url":111},[26,27,28,29,125],"豌豆",{"code":69,"slug":127,"name":128,"path":129,"parent_code":130},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":133,"locked":14,"version":34,"method":35,"evidence":134},"classified",{"candidates":135,"agricultural_evidence":41},[136,144,150],{"code":69,"score":137,"supported":41,"title":138,"abstract":139,"tags":143},12,[],[140,141,142],"深度学习","deep learning","artificial intelligence",[27,28,29],{"code":39,"score":40,"supported":41,"title":145,"abstract":146,"tags":149},[],[147,148],"精准农业","precision agriculture",[26],{"code":47,"score":21,"supported":14,"title":151,"abstract":152,"tags":153},[],[],[125],[155,156],"PEAS-DeiT 豌豆叶病害","DeiT 植物病害分类","PEAS-DeiT豌豆叶病害-4907","10.3390\u002Fapp162010113",{"doi":158,"openalex_id":160,"authors":161,"venue":114,"cited_by_count":96,"oa_url":111,"card":165,"direction":101,"ingested_from":104},"W7221612639",[162],{"name":163,"orcid":164},"Fatma Ozge Ozkok","https:\u002F\u002Forcid.org\u002F0000-0002-1421-4670",{"tldr":166,"method":167,"finding":168,"direction":101,"opportunity":169},"提出PEAS-DeiT模型，将多尺度空间注意力融入DeiT，实现豌豆叶片病害分类。","多尺度空间注意力增强DeiT，构建UPLDD数据集，五折交叉验证与XAI分析。","PEAS-DeiT达到93.11%准确率，优于对比的CNN和Transformer模型。","可探索轻量化Transformer在移动端病害识别中的部署，并扩展多作物跨域泛化能力。","2026-10-14T23:30:46.091813Z",{"id":172,"title":173,"url":174,"summary":175,"summary_zh":176,"content":9,"source_name":177,"source_url":174,"published_at":178,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":179,"sources":181,"tags":183,"primary_topic":9,"secondary_topics":185,"classification":186,"search_phrases":208,"slug":211,"view_count":212,"doi":213,"paper":214,"created_at":229},4437,"A global-local hybrid vision-language framework for interpretable grapevine disease diagnosis","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs44163-026-02463-x","Abstract Plant diseases pose a severe threat to agricultural productivity and global food security. Although Convolutional Neural Networks have achieved impressive classification accuracy in controlled environments, they often lack semantic reasoning capabilities and operate as uninterpretable “black boxes.” Recently, Vision-Language Models have emerged as powerful multimodal reasoning systems. However, they suffer from the “attention sink” phenomenon when applied to complex real-world images, wasting computational resources on background noise rather than microscopic necrotic lesions. In this paper, we propose a Global–Local Hybrid Framework that mimics the diagnostic workflow of agricultural experts. The system leverages an unsupervised Saliency Auto-Crop algorithm within the CIE-LAB color space to automatically extract local disease patches without the need for bounding box annotations. A dual image stream, comprising global context and local patches, is fed into LLaVA−1.5 7B and fine-tuned using 4-bit QLoRA to ensure parameter efficiency. Furthermore, we extract Cross-Attention Heatmaps from the final layer of the decoder to provide Explainable Artificial Intelligence capabilities. Experimental results on a grapevine disease dataset demonstrate that our method achieves an overall accuracy of 98.7%, significantly outperforming zero-shot baselines while providing explicit semantic interpretability for precision agriculture.","摘要 植物病害对农业生产力与全球粮食安全构成严重威胁。尽管卷积神经网络（Convolutional Neural Networks）在受控环境下已取得令人瞩目的分类精度，但其往往缺乏语义推理能力，且作为不可解释的“黑箱”运行。近年来，视觉-语言模型（Vision-Language Models）作为强大的多模态推理系统崭露头角。然而，将其应用于复杂的真实世界图像时，会出现“注意力汇”现象，将计算资源浪费在背景噪声上，而非微小的坏死病斑。本文提出了一种模拟农业专家诊断工作流程的全局-局部混合框架。该系统利用CIE-LAB色彩空间中的无监督显著性自动裁剪算法，无需边界框标注即可自动提取局部病害斑块。由全局上下文和局部斑块组成的双图像流被输入LLaVA−1.5 7B，并使用4位QLoRA进行微调以确保参数效率。此外，我们从解码器最后一层提取交叉注意力热图，以提供可解释人工智能能力。在葡萄藤病害数据集上的实验结果表明，本方法总体准确率达到98.7%，显著优于零样本基线，同时为精准农业提供了明确的语义可解释性。","Discover Artificial Intelligence","2026-10-07T00:00:00Z",{"impact":17,"substance":18,"depth":17,"authority":119,"freshness":120,"relevant":21,"comment":180},"提出全局-局部混合视觉语言框架，以98.7%准确率实现葡萄病害可解释诊断，方法新颖且具产业应用潜力。",[182],{"name":177,"url":174},[26,27,28,30,184],"葡萄病害",[],{"status":33,"locked":14,"version":34,"method":35,"evidence":187},{"candidates":188,"agricultural_evidence":41,"reason":74},[189,193,200,204],{"code":39,"score":40,"supported":41,"title":190,"abstract":191,"tags":192},[],[147,148],[26],{"code":194,"score":48,"supported":41,"title":195,"abstract":196,"tags":199},"A8",[],[197,198],"粮食安全","food security",[],{"code":69,"score":48,"supported":41,"title":201,"abstract":202,"tags":203},[],[142],[27,28,30],{"code":62,"score":70,"supported":41,"title":205,"abstract":206,"tags":207},[],[65],[],[209,210],"葡萄病害 视觉语言模型 诊断","LLaVA QLoRA 葡萄","葡萄病害视觉语言模型诊断-4437",4,"10.1007\u002Fs44163-026-02463-x",{"doi":213,"openalex_id":215,"authors":216,"venue":177,"cited_by_count":96,"oa_url":222,"card":223,"direction":228,"ingested_from":104},"W7220864211",[217,219],{"name":218,"orcid":9},"Lang Hoang Son",{"name":220,"orcid":221},"Bùi Thanh Hùng","https:\u002F\u002Forcid.org\u002F0000-0002-9400-7582","https:\u002F\u002Flink.springer.com\u002Fcontent\u002Fpdf\u002F10.1007\u002Fs44163-026-02463-x.pdf",{"tldr":224,"method":225,"finding":226,"direction":101,"opportunity":227},"提出全局-局部混合视觉语言框架，实现可解释的葡萄病害诊断，准确率达98.7%。","无监督显著性自动裁剪提取局部病斑，双流图像输入LLaVA-1.5 7B，4-bi","方法显著优于零样本基线，并提供跨注意力热图实现可解释诊断。","可探索无标注病斑定位与轻量化VLM在田间实时诊断中的泛化能力。","农业遥感与作物表型","2026-10-08T23:30:29.641451Z",{"id":231,"title":232,"url":233,"summary":234,"summary_zh":9,"content":9,"source_name":235,"source_url":9,"published_at":236,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":237,"score_detail":238,"sources":240,"tags":242,"primary_topic":9,"secondary_topics":244,"classification":245,"search_phrases":263,"slug":266,"view_count":70,"doi":9,"paper":267,"created_at":275},2901,"AgriScope:面向农业图像的像素级多模态理解统一框架,arXiv 2609.20325(预印本)","https:\u002F\u002Farxiv.org\u002Fabs\u002F2609.20325","Boudiaf、Alanssari、Hussain、Javed提出AgriScope,一个统一的像素级多模态农业图像理解框架,联合支持图像级、区域级、像素级理解,可实现接地描述生成、指代表达分割、多轮多模态交互等任务。集成生物专用语义表征、密集空间表征与像素解码;引入大规模像素级农业多模态指令调优数据集AgriGround,包含50万+图像和1100万+指令跟随样本,涵盖植物病害分析、作物与杂草识别、昆虫识别、细粒度植物理解。实验表明AgriScope在多项农业视觉语言任务上有效。","arXiv (preprint)","2026-09-17T00:00:00Z",75,{"impact":17,"substance":18,"depth":17,"authority":20,"freshness":120,"relevant":21,"comment":239},"提出统一像素级农业多模态理解框架并开源50万图像、1100万指令样本的大规模数据集，方法新颖、数据规模突出，但为arXiv预印本、未经同行评审，权威性有限，值得作为前沿技术动态精选。",[241],{"name":235,"url":233},[26,27,243,29,30],"农业遥感",[],{"status":246,"locked":14,"version":34,"method":35,"evidence":247},"unclassified",{"candidates":248,"agricultural_evidence":41,"reason":262},[249,253,257],{"code":69,"score":70,"supported":14,"title":250,"abstract":251,"tags":252},[],[],[27,30,29],{"code":39,"score":21,"supported":14,"title":254,"abstract":255,"tags":256},[],[],[26],{"code":258,"score":21,"supported":14,"title":259,"abstract":260,"tags":261},"I3",[],[],[243],"缺少足够的主题证据",[264,265],"AgriScope 农业图像 多模态","AgriGround 像素级 农业数据集","AgriScope农业图像多模态-2901",{"doi":9,"openalex_id":9,"authors":268,"venue":9,"cited_by_count":96,"oa_url":9,"card":269,"direction":101,"ingested_from":274},[],{"tldr":270,"method":271,"finding":272,"direction":101,"opportunity":273},"提出AgriScope统一框架，实现农业图像像素级多模态理解与多任务交互。","构建AgriGround数据集（50万+图像、1100万+指令样本），融合语义与","AgriScope在接地描述、指代分割、多轮交互等农业视觉语言任务上有效。","可探索像素级多模态模型在田间实时病害诊断与精准施药决策中的落地与轻量化。","agent","2026-09-19T00:06:08.678379Z",{"id":277,"title":278,"url":279,"summary":280,"summary_zh":281,"content":9,"source_name":177,"source_url":279,"published_at":282,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":283,"score_detail":284,"sources":286,"tags":288,"primary_topic":289,"secondary_topics":290,"classification":291,"search_phrases":306,"slug":309,"view_count":96,"doi":310,"paper":311,"created_at":323},2178,"Hybrid explainable DeiT-based framework for plant disease classification and severity estimation","https:\u002F\u002Fdoi.org\u002F10.1007\u002Fs44163-026-02191-2","The detection of plant diseases is essential for preserving agricultural productivity and food security; however, existing approaches often suffer from limited interpretability and generalization capability. This study proposes a hybrid deep learning framework based on Data-efficient Image Transformers (DeiT) for plant disease classification and severity estimation. The framework employs DeiT-Base, DeiT-Small, and DeiT-Tiny models to capture global contextual dependencies in plant leaf images. To improve interpretability, a hybrid Explainable Artificial Intelligence (XAI) module is introduced by combining Gradient-weighted Class Activation Mapping (Grad-CAM) for local feature attribution with Attention Rollout for global dependency visualization. In addition, HSV-based segmentation is applied after classification to isolate disease-relevant regions for damage ratio computation, severity estimation, and explanation refinement. The damage ratio is further integrated with Hybrid XAI attention maps to estimate disease severity. Experiments were conducted on the New Plant Diseases Dataset (Augmented), comprising 70,295 training images and 17,572 validation images across 38 disease classes and 14 plant species. The proposed DeiT-Base model achieved a maximum classification accuracy of 99.13%, outperforming several CNN architectures, including ResNet50, DenseNet121, MobileNetV3, EfficientNet-B4, and InceptionV3. Furthermore, the proposed Hybrid XAI framework demonstrated superior interpretability performance in terms of Focus Score, Background Noise, Signal-to-Noise Ratio (SNR), and Entropy compared with individual explanation methods. Overall, the proposed framework improves classification accuracy, enhances model transparency, and provides meaningful disease severity estimation, making it a promising solution for intelligent precision agriculture.","植物病害检测对于保障农业生产力和粮食安全至关重要，然而现有方法往往存在可解释性有限和泛化能力不足的问题。本研究提出了一种基于数据高效图像Transformer（Data-efficient Image Transformers，DeiT）的混合深度学习框架，用于植物病害分类和严重程度估计。该框架采用DeiT-Base、DeiT-Small和DeiT-Tiny模型来捕获植物叶片图像中的全局上下文依赖关系。为提高可解释性，引入了一种混合可解释人工智能（Explainable Artificial Intelligence，XAI）模块，将用于局部特征归因的梯度加权类激活映射（Gradient-weighted Class Activation Mapping，Grad-CAM）与用于全局依赖可视化的注意力展开（Attention Rollout）相结合。此外，在分类之后应用基于HSV的分割来分离病害相关区域，以进行损伤比率计算、严重程度估计和解释优化。损伤比率进一步与混合XAI注意力图相结合以估计病害严重程度。实验在新植物病害数据集（增强版）（New Plant Diseases Dataset (Augmented)）上进行，该数据集包含70,295张训练图像和17,572张验证图像，涵盖38个病害类别和14种植物物种。所提出的DeiT-Base模型达到了99.13%的最高分类准确率，优于多种CNN架构，包括ResNet50、DenseNet121、MobileNetV3、EfficientNet-B4和InceptionV3。此外，所提出的混合XAI框架在聚焦分数（Focus Score）、背景噪声（Background Noise）、信噪比（Signal-to-Noise Ratio，SNR）和熵（Entropy）方面表现出优于单一解释方法的可解释性性能。总体而言，所提出的框架提高了分类准确率，增强了模型透明度，并提供了有意义的病害严重程度估计，使其成为智能精准农业的一种有前景的解决方案。","2026-09-10T00:00:00Z",78,{"impact":17,"substance":18,"depth":17,"authority":137,"freshness":20,"relevant":21,"comment":285},"基于DeiT与混合可解释AI的植物病害分类与严重度估计研究，在7万张图像、38类病害上取得99.13%准确率，方法新颖、数据扎实，对智慧农业病害智能诊断有参考价值。",[287],{"name":177,"url":279},[26,27,28,147,29],{"code":69,"slug":127,"name":128,"path":129,"parent_code":130},[],{"status":133,"locked":14,"version":34,"method":35,"evidence":292},{"candidates":293,"agricultural_evidence":41},[294,298,302],{"code":69,"score":137,"supported":41,"title":295,"abstract":296,"tags":297},[],[140,141,142],[27,28,29],{"code":39,"score":20,"supported":41,"title":299,"abstract":300,"tags":301},[],[147,148],[26,147],{"code":194,"score":48,"supported":41,"title":303,"abstract":304,"tags":305},[],[197,198],[],[307,308],"农业人工智能 植物病害识别 智慧农业 精准农业","农业人工智能 植物病害识别","农业人工智能植物病害识别智慧农业精准农业-2178","10.1007\u002Fs44163-026-02191-2",{"doi":310,"openalex_id":312,"authors":313,"venue":177,"cited_by_count":96,"oa_url":279,"card":318,"direction":101,"ingested_from":104},"W7212179384",[314,316],{"name":315,"orcid":9},"Muskan Batra",{"name":317,"orcid":9},"Pooja Sharma",{"tldr":319,"method":320,"finding":321,"direction":101,"opportunity":322},"提出基于DeiT的混合可解释框架，实现植物病害分类与严重度估计。","DeiT-Base\u002FSmall\u002FTiny结合Grad-CAM与Attention","DeiT-Base分类准确率达99.13%，混合XAI可解释性指标优于单一方法。","可探索轻量化DeiT在边缘设备部署及多模态数据融合的病害严重度实时估计。","2026-09-11T23:30:47.756846Z",{"id":325,"title":326,"url":327,"summary":328,"summary_zh":329,"content":9,"source_name":330,"source_url":327,"published_at":331,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":332,"score_detail":333,"sources":336,"tags":338,"primary_topic":341,"secondary_topics":347,"classification":349,"search_phrases":366,"slug":368,"view_count":96,"doi":369,"paper":370,"created_at":384},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":20,"substance":334,"depth":17,"authority":137,"freshness":13,"relevant":21,"comment":335},21,"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。",[337],{"name":330,"url":327},[26,27,339,29,340],"小农户","可解释性",{"code":342,"slug":343,"name":344,"path":345,"parent_code":346},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[348],{"code":69,"slug":127,"name":128,"path":129,"parent_code":130},{"status":133,"locked":14,"version":34,"method":35,"evidence":350},{"candidates":351,"agricultural_evidence":41},[352,357,362],{"code":342,"score":137,"supported":41,"title":353,"abstract":355,"tags":356},[354],"smallholder",[354],[339],{"code":69,"score":13,"supported":41,"title":358,"abstract":360,"tags":361},[359],"computer vision",[],[27,29],{"code":39,"score":21,"supported":14,"title":363,"abstract":364,"tags":365},[],[],[26],[367,308],"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可解释性智慧农业-5592","10.5281\u002Fzenodo.23545193",{"doi":369,"openalex_id":371,"authors":372,"venue":330,"cited_by_count":96,"oa_url":327,"card":379,"direction":101,"ingested_from":104},"W7223137323",[373,375,377],{"name":374,"orcid":9},"Akshay Ramgoolam",{"name":376,"orcid":9},"Priya Bundhoo Appadoo",{"name":378,"orcid":9},"Jean-Noël LEspérance",{"tldr":380,"method":381,"finding":382,"direction":101,"opportunity":383},"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":386,"title":387,"url":388,"summary":389,"summary_zh":390,"content":9,"source_name":391,"source_url":388,"published_at":392,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":393,"score_detail":394,"sources":396,"tags":398,"primary_topic":401,"secondary_topics":402,"classification":403,"search_phrases":414,"slug":417,"view_count":96,"doi":418,"paper":419,"created_at":446},5562,"pxGPT: A schema-constrained multimodal LLM framework for plant phenotyping","https:\u002F\u002Fdoi.org\u002F10.1093\u002Fbioinformatics\u002Fbtag781","Abstract Summary Deep learning can scale image-based plant phenotyping but typically targets predefined traits, whereas manual phenotypic description is more flexible but difficult to scale. We present Phenotype eXplorer Generative Plant Traits (pxGPT), a Python framework that converts multi-view plant images into structured phenotype records. The workflow combines free-text description, dataset-specific schema generation with human review, and schema-constrained information extraction. We evaluated the extraction stage on three Brassica datasets using 78,973 annotations from ten annotators who had no access to pxGPT predictions. For 38-45 qualitative traits per dataset, pxGPT-human agreement ranged from 0.572 to 0.712, compared with 0.620 to 0.739 agreement among human annotators. All 32,736 required trait fields were present; 93 values were empty, and no qualitative descriptor violations were recorded. For selected morphological structures, pooled false-positive rates were 0.5-8.0%, while sensitivity varied across structure types in one mature-plant dataset (37.5-100%). The curated schemas were applied unchanged with two commercial models and one locally hosted open-weight model, with differences in mean accuracy of 0.03 or less. In the evaluated Brassica datasets, these results support schema-guided extraction of visible plant traits without separate model training for each trait. Human-guided schema development offers a path to extend pxGPT to other crops and imaging conditions, with accuracy requiring validation in each new setting. Availability and Implementation The software is available at https:\u002F\u002Fgithub.com\u002FYaoChengLab\u002Fpxgpt and an archival version of the source code is available on Zenodo (https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22931745). Curated schemas used in this study, annotation guides, phenotype records and human annotations are available at https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22907554. Supplementary information Supplementary data are available at Bioinformatics online.","摘要 深度学习能够扩展基于图像的植物表型分析，但通常针对预定义性状，而人工表型描述更为灵活却难以规模化。我们提出Phenotype eXplorer Generative Plant Traits（pxGPT），一个将多视角植物图像转换为结构化表型记录的Python框架。该工作流程结合了自由文本描述、数据集特定模式生成与人工审核，以及模式约束的信息提取。我们在三个芸薹属（Brassica）数据集上评估了提取阶段，使用了来自十名标注者的78,973条标注，这些标注者无法获取pxGPT的预测结果。对于每个数据集38-45个定性性状，pxGPT与人工标注者之间的一致性范围为0.572至0.712，而人工标注者之间的一致性为0.620至0.739。所有32,736个所需性状字段均已存在；93个值为空，未记录到定性描述符违规。对于选定的形态结构，合并假阳性率为0.5-8.0%，而在一个成熟植株数据集中，灵敏度因结构类型而异（37.5-100%）。经整理的模式在未作修改的情况下应用于两个商业模型和一个本地部署的开放权重模型，平均准确率差异不超过0.03。在所评估的芸薹属数据集中，这些结果支持在无需为每个性状单独训练模型的情况下，对可见植物性状进行模式引导的提取。人工引导的模式开发为将pxGPT扩展至其他作物和成像条件提供了途径，其准确性需在每个新设置中加以验证。可用性与实现 该软件可在https:\u002F\u002Fgithub.com\u002FYaoChengLab\u002Fpxgpt获取，源代码的存档版本可在Zenodo（https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22931745）获取。本研究中使用的经整理模式、标注指南、表型记录和人工标注可在https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.22907554获取。补充信息 补充数据可在Bioinformatics在线获取。","Bioinformatics","2026-10-20T00:00:00Z",81,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":120,"relevant":21,"comment":395},"提出基于模式约束的多模态大模型植物表型提取框架，在近8万条标注上验证，方法新颖、数据规模可观，对智慧育种与表型自动化有参考价值。",[397],{"name":391,"url":388},[26,27,399,30,400],"植物表型","芸薹属",{"code":69,"slug":127,"name":128,"path":129,"parent_code":130},[],{"status":133,"locked":14,"version":34,"method":35,"evidence":404},{"candidates":405,"agricultural_evidence":41},[406,410],{"code":69,"score":20,"supported":41,"title":407,"abstract":408,"tags":409},[],[140,141],[27,30],{"code":39,"score":21,"supported":14,"title":411,"abstract":412,"tags":413},[],[],[26],[415,416],"pxGPT 植物表型","Brassica 表型 多模态大模型","pxGPT植物表型-5562","10.1093\u002Fbioinformatics\u002Fbtag781",{"doi":418,"openalex_id":420,"authors":421,"venue":391,"cited_by_count":96,"oa_url":440,"card":441,"direction":103,"ingested_from":104},"W7223127960",[422,425,427,430,433,436,438],{"name":423,"orcid":424},"Po‐Xing Zheng","https:\u002F\u002Forcid.org\u002F0000-0001-5827-5865",{"name":426,"orcid":9},"Jheng-Yan Ou",{"name":428,"orcid":429},"Yu-Lin Wu","https:\u002F\u002Forcid.org\u002F0009-0003-8596-9047",{"name":431,"orcid":432},"Ruei-Shiuan Wang","https:\u002F\u002Forcid.org\u002F0009-0007-5159-1129",{"name":434,"orcid":435},"WU Ting-li","https:\u002F\u002Forcid.org\u002F0000-0001-7996-5515",{"name":437,"orcid":9},"Te-Chang Hsu",{"name":439,"orcid":9},"Yao-Cheng Lin","https:\u002F\u002Facademic.oup.com\u002Fbioinformatics\u002Fadvance-article-pdf\u002Fdoi\u002F10.1093\u002Fbioinformatics\u002Fbtag781\u002F72201337\u002Fbtag781.pdf",{"tldr":442,"method":443,"finding":444,"direction":228,"opportunity":445},"提出pxGPT框架，用多模态大模型将多视角植物图像转为结构化表型记录。","结合自由文本描述、人工审核的模式生成与模式约束信息抽取，在三个芸薹属数据集上评估","pxGPT与人工标注一致性0.572-0.712，接近人工间一致性，且无需逐性状训练模型。","可探索将模式引导抽取迁移至更多作物与复杂成像条件，并研究自动化模式生成以降低人工审核成本。","2026-10-22T11:20:22.173561Z"]