[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-4042":3,"related-4042":87},{"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":22,"tags":24,"primary_topic":30,"secondary_topics":36,"classification":37,"search_phrases":65,"slug":68,"view_count":20,"doi":69,"paper":70,"created_at":86},4042,"Comparative Analysis of Decision-Level Ensemble Voting Strategies for Tomato Leaf Disease Classification Using EfficientNet-B0 and MobileNetV2","https:\u002F\u002Fdoi.org\u002F10.15294\u002Frji.v4i2.61961","Abstract. Tomato leaf diseases are a major threat to agricultural productivity, making accurate and timely disease classification essential for supporting precision agriculture. Recent advances in convolutional neural networks have significantly improved automated plant disease recognition, while ensemble learning has further enhanced classification performance by combining predictions from multiple models. Among various ensemble approaches, decision-level voting strategies, including hard voting, soft voting, and weighted soft voting, are widely adopted because of their simplicity and effectiveness. However, systematic comparisons of these voting strategies, particularly when combining lightweight CNN architectures for tomato leaf disease classification, remain limited. Therefore, this study investigates the comparative performance of different decision-level voting strategies to provide empirical evidence for selecting the most appropriate ensemble approach. Purpose: This study aims to systematically compare hard voting, soft voting, and weighted soft voting for combining EfficientNet-B0 and MobileNetV2 to identify the most effective decision-level ensemble strategy for tomato leaf disease classification. Methods\u002FStudy design\u002Fapproach: EfficientNet-B0 and MobileNetV2 were trained independently and combined using hard voting with confidence-based tie breaking, soft voting, and weighted soft voting, with voting weights optimized through grid search. Performance was evaluated using accuracy, macro-F1 score, confusion matrices, and McNemar's test. Result\u002FFindings: EfficientNet-B0 outperformed MobileNetV2 as the best single model (94.05% accuracy, 0.9266 macro-F1 vs. 90.64% accuracy, 0.8781 macro-F1). All three ensemble strategies further improved classification performance, with weighted soft voting achieving the highest performance (95.15% accuracy, 0.9417 macro-F1). McNemar's test showed that only soft voting and weighted soft voting significantly outperformed the best single model, whereas hard voting did not. No statistically significant differences were observed among the three ensemble voting strategies. Novelty: This study systematically compares hard voting, soft voting, and weighted soft voting for decision-level ensemble using EfficientNet-B0 and MobileNetV2 in tomato leaf disease classification. In addition to evaluating predictive performance, the study incorporates McNemar's statistical test to determine whether the observed performance differences among the ensemble strategies are statistically significant, providing empirical guidance for selecting appropriate voting strategies.","摘要。番茄叶片病害是农业生产力的主要威胁，因此准确、及时地进行病害分类对于支持精准农业至关重要。近年来，卷积神经网络的进展显著提升了植物病害的自动识别能力，而集成学习通过结合多个模型的预测结果进一步提高了分类性能。在各种集成方法中，决策级投票策略，包括硬投票、软投票和加权软投票，因其简单有效而被广泛采用。然而，针对这些投票策略的系统比较，尤其是在结合轻量级CNN架构进行番茄叶片病害分类时，仍然有限。因此，本研究考察了不同决策级投票策略的比较性能，为选择最合适的集成方法提供实证依据。目的：本研究旨在系统比较硬投票、软投票和加权软投票在结合EfficientNet-B0和MobileNetV2时的表现，以确定番茄叶片病害分类中最有效的决策级集成策略。方法\u002F研究设计\u002F研究路径：EfficientNet-B0和MobileNetV2分别独立训练，并通过基于置信度的平局打破硬投票、软投票和加权软投票进行结合，其中投票权重通过网格搜索进行优化。采用准确率、宏F1分数、混淆矩阵和McNemar检验评估性能。结果\u002F发现：EfficientNet-B0作为最佳单一模型优于MobileNetV2（准确率94.05%、宏F1分数0.9266，对比准确率90.64%、宏F1分数0.8781）。三种集成策略均进一步提升了分类性能，其中加权软投票表现最佳（准确率95.15%、宏F1分数0.9417）。McNemar检验表明，只有软投票和加权软投票显著优于最佳单一模型，而硬投票则没有。三种集成投票策略之间未观察到统计学显著差异。创新性：本研究系统比较了硬投票、软投票和加权软投票在番茄叶片病害分类中使用EfficientNet-B0和MobileNetV2进行决策级集成的表现。除评估预测性能外，本研究还引入McNemar统计检验，以确定所观察到的性能差异",null,"Recursive Journal of Informatics","2026-09-30T00:00:00Z","论文",10,false,62,{"impact":17,"substance":18,"depth":19,"authority":13,"freshness":17,"relevant":20,"comment":21},8,20,16,1,"方法对比扎实、含统计检验与量化结论，但属细分模型研究，产业影响有限，适合作为技术类精选。",[23],{"name":10,"url":6},[25,26,27,28,29],"智慧农业","农业人工智能","集成学习","番茄种植","病害识别",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},"B1","vegetables","蔬菜","\u002Ftopic\u002Fhorticulture\u002Fvegetables","B",[],{"status":38,"locked":14,"version":39,"method":40,"evidence":41},"classified","2026-10-22.2","rules",{"candidates":42,"agricultural_evidence":45},[43,51,59],{"code":31,"score":44,"supported":45,"title":46,"abstract":48,"tags":50},15,true,[47],"tomato",[49,47],"番茄",[28],{"code":52,"score":53,"supported":45,"title":54,"abstract":55,"tags":58},"I1",7,[],[56,57],"精准农业","precision agriculture",[25],{"code":60,"score":61,"supported":14,"title":62,"abstract":63,"tags":64},"I2",2,[],[],[26,29],[66,67],"EfficientNet-B0 MobileNetV2 番茄叶病","番茄叶病 集成投票 分类","EfficientNet-B0MobileNetV2番茄叶病-4042","10.15294\u002Frji.v4i2.61961",{"doi":69,"openalex_id":71,"authors":72,"venue":10,"cited_by_count":78,"oa_url":9,"card":79,"direction":83,"ingested_from":85},"W7215044289",[73,75],{"name":74,"orcid":9},"Hani Nafis Muzaki",{"name":76,"orcid":77},"Yahya Nur Ifriza","https:\u002F\u002Forcid.org\u002F0000-0003-1109-1959",0,{"tldr":80,"method":81,"finding":82,"direction":83,"opportunity":84},"系统比较硬投票、软投票与加权软投票三种决策级集成策略，用于番茄叶片病害分类。","EfficientNet-B0与MobileNetV2独立训练后集成，用网格搜索","加权软投票最优（准确率95.15%，宏F1 0.9417），仅软投票与加权软投票显著优于最佳单模型。","农业人工智能与决策模型","可探索更多轻量模型组合与动态权重策略，并在真实田间复杂背景下验证集成方法的鲁棒性。","openalex","2026-10-02T23:30:51.207730Z",{"total":88,"page":20,"page_size":88,"items":89},6,[90,152,202,288,342,405],{"id":91,"title":92,"url":93,"summary":94,"summary_zh":9,"content":9,"source_name":95,"source_url":9,"published_at":96,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":97,"score_detail":98,"sources":103,"tags":105,"primary_topic":107,"secondary_topics":108,"classification":109,"search_phrases":138,"slug":141,"view_count":61,"doi":142,"paper":143,"created_at":151},5301,"MambaSeed:基于Mamba的轻量级混合视觉网络用于番茄叶病识别(PLOS ONE)","https:\u002F\u002Fjournals.plos.org\u002Fplosone\u002Farticle?id=10.1371\u002Fjournal.pone.0359350","PLOS ONE于2026-10-15发表论文MambaSeed,由Wang X, Zhang H, Chen Y, Li Y, Qin Y, Lv Z等团队提出基于MambaVisionMixer的轻量级混合视觉网络,配套局部连续分支注入(LCBI)机制;在番茄、玉米、马铃薯数据集上识别准确率分别达98.97%、98.06%和95.87%,模型仅1.13M参数、0.240G MACs,在RTX 4090上延迟3.39ms\u002F图像,为资源受限农田环境提供紧凑、高效、鲁棒的病害识别方案。","PLOS ONE(2026-10-15)","2026-10-15T00:00:00Z",79,{"impact":99,"substance":100,"depth":99,"authority":101,"freshness":17,"relevant":20,"comment":102},18,22,13,"方法新颖、指标扎实的轻量级病害识别论文，对资源受限农田场景有实用价值，值得进入每日精选。",[104],{"name":95,"url":93},[25,26,49,29,106],"轻量级模型",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[],{"status":38,"locked":14,"version":39,"method":40,"evidence":110},{"candidates":111,"agricultural_evidence":45},[112,117,124,130,134],{"code":31,"score":113,"supported":45,"title":114,"abstract":115,"tags":116},12,[49],[49],[49],{"code":118,"score":119,"supported":45,"title":120,"abstract":121,"tags":123},"A3",3,[],[122],"玉米",[],{"code":125,"score":119,"supported":45,"title":126,"abstract":127,"tags":129},"A5",[],[128],"马铃薯",[],{"code":60,"score":61,"supported":14,"title":131,"abstract":132,"tags":133},[],[],[26,29],{"code":52,"score":20,"supported":14,"title":135,"abstract":136,"tags":137},[],[],[25],[139,140],"MambaSeed 番茄叶病","MambaVisionMixer 病害识别","MambaSeed番茄叶病-5301","10.1371\u002Fjournal.pone.0359350",{"doi":142,"openalex_id":9,"authors":144,"venue":9,"cited_by_count":78,"oa_url":9,"card":145,"direction":83,"ingested_from":150},[],{"tldr":146,"method":147,"finding":148,"direction":83,"opportunity":149},"提出轻量级混合视觉网络MambaSeed，实现番茄等作物叶病高效识别。","基于MambaVisionMixer与局部连续分支注入机制，在三种作物病害数据集","番茄、玉米、马铃薯识别准确率达98.97%、98.06%、95.87%，仅1.13M参数、3.39m","可探索该轻量模型在移动端与边缘设备的田间部署，及多作物跨域泛化与病害早期识别。","agent","2026-10-19T00:07:48.589614Z",{"id":153,"title":154,"url":155,"summary":156,"summary_zh":9,"content":9,"source_name":157,"source_url":9,"published_at":158,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":159,"score_detail":160,"sources":163,"tags":165,"primary_topic":167,"secondary_topics":168,"classification":169,"search_phrases":190,"slug":193,"view_count":61,"doi":9,"paper":194,"created_at":201},2904,"Decoupled Foundation Models:基于YOLO26m+SAM2+DINOv2的湿度诱导番茄叶坏死实例分割与检测,登MDPI Agriculture 16(18)1997","https:\u002F\u002Fwww.mdpi.com\u002F2077-0472\u002F16\u002F18\u002F1997","本研究针对温室番茄相对湿度过高引发的非生物胁迫(生理性叶坏死,与生物感染症状相似),提出多步AI管道自动化分割与分类坏死叶斑。采集218张RGB图像、3218个标注(棕色坏死斑\u002F黄色坏死斑\u002F无坏死),系统评估6种端到端实例分割管道(YOLO26m检测+SAM2零样本分割+微调DINOv2或EfficientNet-B3分类);微调DINOv2宏F1达0.926,优于EfficientNet-B3、ResNet-50、Swin-Small基线(0.886-0.901);最佳配置mAP@50=0.828,较YOLO26m单模型提升约8%。","MDPI Agriculture","2026-09-17T00:00:00Z",78,{"impact":19,"substance":100,"depth":99,"authority":101,"freshness":161,"relevant":20,"comment":162},9,"方法组合新颖、数据规模与对比基线扎实，对温室番茄生理性叶坏死自动识别有实用价值，值得进入每日精选。",[164],{"name":157,"url":155},[25,26,166,49,29],"设施农业",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[],{"status":38,"locked":14,"version":39,"method":40,"evidence":170},{"candidates":171,"agricultural_evidence":45},[172,176,182,186],{"code":31,"score":113,"supported":45,"title":173,"abstract":174,"tags":175},[49],[49],[49],{"code":177,"score":119,"supported":45,"title":178,"abstract":179,"tags":181},"B7",[],[180],"温室",[],{"code":60,"score":61,"supported":14,"title":183,"abstract":184,"tags":185},[],[],[26,29],{"code":52,"score":20,"supported":14,"title":187,"abstract":188,"tags":189},[],[],[25],[191,192],"番茄叶坏死 实例分割","农业人工智能 智慧农业 病害识别 设施农业","番茄叶坏死实例分割-2904",{"doi":9,"openalex_id":9,"authors":195,"venue":9,"cited_by_count":78,"oa_url":9,"card":196,"direction":83,"ingested_from":150},[],{"tldr":197,"method":198,"finding":199,"direction":83,"opportunity":200},"用YOLO26m+SAM2+DINOv2多步管道分割并分类高湿诱导的番茄叶坏死斑。","218张RGB图像、3218个标注，评估6种实例分割管道并微调DINOv2分类。","微调DINOv2宏F1达0.926，最佳配置mAP@50=0.828，较单模型提升约8%。","可探索零样本基础模型在多种非生物胁迫症状上的泛化与轻量化温室部署。","2026-09-19T00:06:09.021594Z",{"id":203,"title":204,"url":205,"summary":206,"summary_zh":207,"content":9,"source_name":208,"source_url":205,"published_at":158,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":209,"score_detail":210,"sources":215,"tags":217,"primary_topic":220,"secondary_topics":221,"classification":232,"search_phrases":261,"slug":264,"view_count":61,"doi":265,"paper":266,"created_at":287},2805,"Synergistic effects of spectral preprocessing and machine learning algorithms for nitrogen estimation in tomato using hyperspectral spectral data","https:\u002F\u002Fdoi.org\u002F10.1038\u002Fs41598-026-71908-1","Abstract Accurate estimation of leaf nitrogen content is essential for optimizing fertilization management and improving crop productivity. This study proposes a nondestructive hyperspectral imaging framework combined with advanced machine learning algorithms to classify nitrogen levels in tomato leaves (Solanum lycopersicum L., Royal variety). A total of 300 hyperspectral samples were collected from plants subjected to three nitrogen treatments (N-30%, N-60%, and N-90%) under controlled conditions. Reflectance data (400–1000 nm) were calibrated and preprocessed using Standard Normal Variate (SNV), Multiplicative Scatter Correction (MSC), and Savitzky–Golay (SG) filtering. Both unsupervised and supervised algorithms were systematically evaluated. Clustering results demonstrated that preprocessing substantially influenced class separability. The MSC + GMM combination yielded the best unsupervised results, with the lowest Davies–Bouldin index (0.443) and the highest silhouette coefficient (0.975), indicating improved cluster compactness and separation. In supervised learning, Neural Networks (NN) consistently outperformed the other models, achieving 100% accuracy and an ROC–AUC of 1.00 under SNV, MSC, and SG preprocessing. Gradient Boosting (GB) and Support Vector Machine (SVM) also demonstrated strong predictive capability, whereas conventional models, including LDA, LR, KNN, and NB, showed moderate improvements following preprocessing. Statistical analysis confirmed the significant effect of spectral preprocessing on both clustering and classification outcomes ( p \u003C 0.05). Overall, integrating hyperspectral imaging with appropriate preprocessing and nonlinear machine learning models provided strong classification performance under the controlled experimental conditions. However, the exceptionally high performance observed for some models should be interpreted cautiously because of the relatively small dataset and experimentally controlled nitrogen treatments. Moreover, the absence of an independent external validation dataset limits the assessment of generalizability across growing conditions, cultivars, and field environments. Therefore, independent validation using larger and more diverse datasets is required before broader application of the proposed framework in practical precision agriculture.","摘要 准确估算叶片氮含量对于优化施肥管理和提高作物生产力至关重要。本研究提出了一种无损高光谱成像框架，结合先进的机器学习算法对番茄叶片（Solanum lycopersicum L.，Royal品种）的氮水平进行分类。在受控条件下，从接受三种氮处理（N-30%、N-60%和N-90%）的植株中共采集了300个高光谱样本。反射率数据（400–1000 nm）经标准正态变量变换（SNV）、多元散射校正（MSC）和Savitzky–Golay（SG）滤波进行校准和预处理。系统评估了无监督和有监督算法。聚类结果表明，预处理对类别可分性产生了显著影响。MSC + GMM组合取得了最佳无监督结果，具有最低的Davies–Bouldin指数（0.443）和最高的轮廓系数（0.975），表明聚类紧密度和分离度得到改善。在有监督学习中，神经网络（NN）始终优于其他模型，在SNV、MSC和SG预处理下均达到100%的准确率和1.00的ROC–AUC。梯度提升（GB）和支持向量机（SVM）也表现出较强的预测能力，而传统模型（包括LDA、LR、KNN和NB）在预处理后表现出中等程度的改善。统计分析证实了光谱预处理对聚类和分类结果均有显著影响（p \u003C 0.05）。总体而言，将高光谱成像与适当的预处理及非线性机器学习模型相结合，在受控实验条件下提供了较强的分类性能。然而，由于数据集相对较小且氮处理为实验受控条件，某些模型所表现出的极高性能应谨慎解读。此外，缺乏独立的外部验证数据集限制了对跨生长条件、品种和田间环境泛化能力的评估。因此，在将该框架广泛应用于实际精准农业之前，需要使用更大规模、更多样化的数据集进行独立验证。","Scientific Reports",77,{"impact":44,"substance":211,"depth":212,"authority":213,"freshness":13,"relevant":20,"comment":214},21,17,14,"高光谱结合机器学习实现番茄叶片氮素无损估测，方法系统、结论明确，但样本量小且缺乏外部验证，属细分领域方法学进展。",[216],{"name":208,"url":205},[25,26,218,28,219],"精准施肥","高光谱遥感",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[222,227],{"code":60,"slug":223,"name":224,"path":225,"parent_code":226},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",{"code":228,"slug":229,"name":230,"path":231,"parent_code":226},"I3","remote-sensing","遥感与农情监测","\u002Ftopic\u002Fdigital-agriculture\u002Fremote-sensing",{"status":38,"locked":14,"version":39,"method":40,"evidence":233},{"candidates":234,"agricultural_evidence":45},[235,239,245,250,257],{"code":31,"score":44,"supported":45,"title":236,"abstract":237,"tags":238},[47],[49,47],[28],{"code":60,"score":44,"supported":45,"title":240,"abstract":242,"tags":244},[241],"machine learning",[243,241],"机器学习",[26],{"code":228,"score":113,"supported":45,"title":246,"abstract":248,"tags":249},[247],"hyperspectral",[247],[219],{"code":251,"score":53,"supported":45,"title":252,"abstract":253,"tags":256},"F3",[],[254,255],"施肥","fertilization",[218],{"code":52,"score":53,"supported":45,"title":258,"abstract":259,"tags":260},[],[56,57],[25],[262,263],"农业人工智能 高光谱遥感 智慧农业 番茄种植","农业人工智能 高光谱遥感","农业人工智能高光谱遥感智慧农业番茄种植-2805","10.1038\u002Fs41598-026-71908-1",{"doi":265,"openalex_id":267,"authors":268,"venue":208,"cited_by_count":78,"oa_url":205,"card":281,"direction":83,"ingested_from":85},"W7213472002",[269,272,275,278],{"name":270,"orcid":271},"Mohammad Vahedi Torshizi","https:\u002F\u002Forcid.org\u002F0000-0003-3648-1515",{"name":273,"orcid":274},"Sajad Sabzi","https:\u002F\u002Forcid.org\u002F0000-0003-2439-5329",{"name":276,"orcid":277},"Mohsen Azadbakht","https:\u002F\u002Forcid.org\u002F0000-0002-5726-9321",{"name":279,"orcid":280},"Razieh Pourdarbani","https:\u002F\u002Forcid.org\u002F0000-0003-0766-8305",{"tldr":282,"method":283,"finding":284,"direction":285,"opportunity":286},"用高光谱成像结合预处理与机器学习，对番茄叶片氮素水平进行分类。","300个高光谱样本，SNV、MSC、SG预处理，聚类与多种监督分类算法对比。","MSC+GMM聚类最优，神经网络在三种预处理下均达100%准确率。","农业遥感与作物表型","样本少且无外部验证，可扩展多品种、多环境田间数据并做独立验证以提升泛化性。","2026-09-17T23:30:59.361085Z",{"id":289,"title":290,"url":291,"summary":292,"summary_zh":293,"content":9,"source_name":294,"source_url":291,"published_at":295,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":296,"score_detail":297,"sources":299,"tags":301,"primary_topic":304,"secondary_topics":305,"classification":307,"search_phrases":324,"slug":327,"view_count":119,"doi":328,"paper":329,"created_at":341},2158,"EfficientNet-CBAM-prototype: an attention-guided EfficientNet framework with dynamic prototype representation for tomato disease classification","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffpls.2026.1918175","Introduction In the field of precision agriculture, one of the major hurdles is the early and accurate identification of plant diseases. Farmers may face serious irreversible loss in yield if there is a delay in diagnosis by even a few days. The CNN model has helped in improving the classification of plant diseases but faces difficulty in distinguishing fine-grained symptoms, has limited generalizability, and is not easily interpretable. Though the deep CNN model performs well in classifying plant diseases, the current methods have several flaws. Most importantly, existing algorithms misclassify visually complex samples due to their limited spatial differentiation of disease-specific morphological features, including lesion borders and necrotic regions. Classification reliability is further compromised by low inter-class embedding separability for visually comparable illness phenotypes. Apart from these representational problems, training instability is still a major problem for attention-based models. Combining randomly initialized attention modules with pretrained backbone networks causes this instability, which still limits practical application. Methods We suggest EfficientNet-CBAM-Prototype (ECP-Net), a unique end-to-end deep learning architecture, to overcome these constraints. Three complementary techniques are combined into a single framework by ECP-Net. First, parameter-efficient multi-scale feature extraction is done using an EfficientNetB0 backbone. Second, joint channel-wise and spatial feature recalibration is performed using a stabilized convolutional block attention module (CBAM). Third, a dynamic prototype memory layer uses cosine similarity-based categorization and exponential moving average (EMA) updates to maintain class-representative embedding vectors. To overcome the instability caused by randomly initialized attention weights, we further propose a two-phase training strategy wherein CBAM is frozen during phase 1 to allow prototype stabilization and then jointly fine-tuned with the learning rate in phase 2. Results Evaluated on the PlantVillage tomato subset comprising 10 disease classes across a class-balanced split of 10,000 training, 500 validation, and 500 test samples, ECP-Net achieves 98.6% test accuracy, 98.59% F1-score, and 98.65% precision with only 4.80M parameters and 85.04 ms average inference time. These results outperformed baselines including VGG16 (97.00%), ResNet50 (81.20%), MobileNetV2 (81.20%), and CNN (70.00%). Discussion Generalization is further validated on 35 real-field tomato leaf images captured under natural, uncontrolled conditions, confirming practical deployment potential.","引言 在精准农业领域，植物病害的早期准确识别是主要难题之一。若诊断延迟哪怕几天，农民可能面临严重的不可逆产量损失。CNN模型有助于改进植物病害分类，但在区分细粒度症状方面存在困难，泛化能力有限，且不易解释。尽管深度CNN模型在植物病害分类中表现良好，当前方法仍存在若干缺陷。最重要的是，现有算法由于对病害特异性形态特征（包括病斑边缘和坏死区域）的空间区分能力有限，会对视觉上复杂的样本产生误分类。对于视觉上相似的病害表型，类间嵌入可分性低，进一步损害了分类可靠性。除这些表征问题外，训练不稳定性仍是基于注意力模型的主要问题。随机初始化的注意力模块与预训练骨干网络相结合会导致这种不稳定性，这仍然限制了实际应用。方法 我们提出EfficientNet-CBAM-Prototype（ECP-Net），一种新颖的端到端深度学习架构，以克服这些限制。ECP-Net将三种互补技术结合到一个统一框架中。首先，使用EfficientNetB0骨干网络进行参数高效的多尺度特征提取。其次，使用稳定化的卷积块注意力模块（CBAM）进行通道与空间的联合特征重校准。第三，动态原型记忆层使用基于余弦相似度的分类和指数移动平均（EMA）更新来维护类代表性嵌入向量。为克服随机初始化注意力权重引起的不稳定性，我们进一步提出两阶段训练策略，其中CBAM在第一阶段冻结以使原型稳定，然后在第二阶段以学习率联合微调。结果 在PlantVillage番茄子集上评估，该子集包含10个病害类别，按类别均衡划分的10,000个训练样本、500个验证样本和500个测试样本，ECP-Net达到98.6%的测试准确率、98.59%的F1分数和98.65%的精确率，仅需4.80M参数和85.04 ms平均推理时间。这些结果优于基线，包括VGG16（97.00%）、ResNet50（81.20%）、MobileNetV2（81.20%）和CNN（70.00%）。","Frontiers in Plant Science","2026-09-10T00:00:00Z",81,{"impact":99,"substance":100,"depth":99,"authority":213,"freshness":161,"relevant":20,"comment":298},"提出注意力引导与动态原型结合的番茄病害识别框架，在公开数据集上取得98.6%准确率并完成真实田间图像验证，方法新颖、数据扎实，对作物病害智能诊断有实用参考价值。",[300],{"name":294,"url":291},[25,26,302,303,28],"深度学习","病虫害识别",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[306],{"code":60,"slug":223,"name":224,"path":225,"parent_code":226},{"status":38,"locked":14,"version":39,"method":40,"evidence":308},{"candidates":309,"agricultural_evidence":45},[310,315,320],{"code":31,"score":99,"supported":45,"title":311,"abstract":312,"tags":314},[47],[49,313,47],"茄子",[28],{"code":60,"score":161,"supported":45,"title":316,"abstract":317,"tags":319},[],[302,318],"deep learning",[26,302,303],{"code":52,"score":53,"supported":45,"title":321,"abstract":322,"tags":323},[],[56,57],[25],[325,326],"农业人工智能 病虫害识别 智慧农业 深度学习","农业人工智能 病虫害识别","农业人工智能病虫害识别智慧农业深度学习-2158","10.3389\u002Ffpls.2026.1918175",{"doi":328,"openalex_id":330,"authors":331,"venue":294,"cited_by_count":78,"oa_url":291,"card":336,"direction":285,"ingested_from":85},"W7212120374",[332,334],{"name":333,"orcid":9},"E Jansi",{"name":335,"orcid":9},"Kavitha BR",{"tldr":337,"method":338,"finding":339,"direction":83,"opportunity":340},"提出ECP-Net，用注意力与动态原型提升番茄病害分类精度与可解释性。","EfficientNetB0+稳定CBAM+动态原型记忆层，两阶段训练策略。","在PlantVillage番茄10类上达98.6%准确率，仅4.80M参数，优于多个基线。","可探索原型可解释性在田间复杂背景与跨域病害识别中的泛化能力。","2026-09-11T23:30:25.043704Z",{"id":343,"title":344,"url":345,"summary":346,"summary_zh":347,"content":9,"source_name":348,"source_url":345,"published_at":349,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":350,"score_detail":351,"sources":353,"tags":355,"primary_topic":356,"secondary_topics":357,"classification":363,"search_phrases":387,"slug":389,"view_count":61,"doi":390,"paper":391,"created_at":404},2048,"Optimized Inception-v4 CNN Combined with YOLOv8 andXGBoost for Tomato Plant Disease Recognition and Classification","https:\u002F\u002Fdoi.org\u002F10.38094\u002Fjastt711230","This study proposes an integrated hybrid pipeline approach which is the combination of an optimized Inception-v4 Convolutional Neural Network (CNN), YOLOv8 (You Only Look Once version 8) and extreme gradient boosting (XG Boost) models to detect and classify Tomato plant diseases effectively and accurately. The proposed methodology is based on the fine-grained localization ability of the YOLOv8 model to accurately localize the affected area of the leaf, multi-scale deep feature extraction ability of the optimized Inceptionv4 CNN, and XG Boost to reduce dimensions and optimize features. The hybrid model combines the Inception-v4 CNN, YOLOv8, and XG Boost models with an average CPU processing time of about 200 milliseconds per image, the model performs better in terms of computational efficiency, robustness, and faster prediction. Experimental tests on Plant Village show that the model outperforms the state of the art in 10 disease categories. Specifically, the integrated hybrid system achieved high precision, recall, F1-score, mean Average Precision (m AP), and its training accuracy was more than 0.98 (up to 0.9969), and the testing accuracy was more than 0.96 (up to 0.9695). The hybrid framework shows good reliability even for diseases that are visually similar, such as bacterial spot, early blight, late blight, mosaic virus and yellow leaf curl virus. The optimization of batch size to 32, along with the tuning of the learning rate further improved the stability of training, convergence speed and the generalization of the overall model. The proposed system holds promise for real-time, scalable, and sustainable precision agriculture, aiming for early disease detection and yield protection. Future work includes incorporating Internet of Things (IoT) edge devices, in field environmental monitoring, multimodal data integration and applying Explainable Artificial Intelligence (EAI) techniques to enhance model interpretability for end-users.","本研究提出了一种集成的混合流水线方法，将优化的Inception-v4卷积神经网络（CNN）、YOLOv8（You Only Look Once version 8）和极端梯度提升（XG Boost）模型相结合，以有效且准确地检测和分类番茄植株病害。所提方法基于YOLOv8模型的细粒度定位能力来准确定位叶片受感染区域、优化的Inception-v4 CNN的多尺度深层特征提取能力，以及XG Boost用于降维和特征优化。该混合模型将Inception-v4 CNN、YOLOv8和XG Boost模型结合在一起，每幅图像的平均CPU处理时间约为200毫秒，模型在计算效率、鲁棒性和更快预测方面表现更优。在Plant Village上的实验测试表明，该模型在10个病害类别上优于现有最先进方法。具体而言，该集成混合系统实现了较高的精确率、召回率、F1分数、平均精度均值（mAP），其训练准确率超过0.98（最高达0.9969），测试准确率超过0.96（最高达0.9695）。即使对于视觉上相似的病害，如细菌性斑点病、早疫病、晚疫病、花叶病毒和黄花叶卷曲病毒，该混合框架也表现出良好的可靠性。将批量大小优化为32，并调整学习率，进一步提高了训练稳定性、收敛速度和整体模型的泛化能力。所提系统有望用于实时、可扩展和可持续的精准农业，旨在实现早期病害检测和产量保护。未来工作包括整合物联网（IoT）边缘设备、田间环境监测、多模态数据融合，以及应用可解释人工智能（EAI）技术以增强模型对最终用户的可解释性。","Journal of Applied Science and Technology Trends","2026-09-09T00:00:00Z",75,{"impact":19,"substance":18,"depth":99,"authority":113,"freshness":161,"relevant":20,"comment":352},"提出YOLOv8+Inception-v4+XGBoost混合模型用于番茄病害识别，准确率与效率数据扎实，对农业AI病害检测有参考价值。",[354],{"name":348,"url":345},[25,26,303,56,28],{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[358],{"code":359,"slug":360,"name":361,"path":362,"parent_code":226},"I5","agricultural-iot","物联网传感与边缘计算","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-iot",{"status":38,"locked":14,"version":39,"method":40,"evidence":364},{"candidates":365,"agricultural_evidence":45},[366,370,377,381],{"code":31,"score":44,"supported":45,"title":367,"abstract":368,"tags":369},[47],[49,47],[28],{"code":359,"score":161,"supported":45,"title":371,"abstract":372,"tags":376},[],[373,374,375],"物联网","internet of things","IoT",[],{"code":52,"score":17,"supported":45,"title":378,"abstract":379,"tags":380},[],[56,57],[25,56],{"code":60,"score":382,"supported":45,"title":383,"abstract":384,"tags":386},5,[],[385],"artificial intelligence",[26,303],[388,326],"农业人工智能 病虫害识别 智慧农业 番茄种植","农业人工智能病虫害识别智慧农业番茄种植-2048","10.38094\u002Fjastt711230",{"doi":390,"openalex_id":392,"authors":393,"venue":348,"cited_by_count":78,"oa_url":345,"card":398,"direction":403,"ingested_from":85},"W7212078878",[394,396],{"name":395,"orcid":9},"Sowmya B",{"name":397,"orcid":9},"Guruprasad S",{"tldr":399,"method":400,"finding":401,"direction":83,"opportunity":402},"提出YOLOv8+Inception-v4+XGBoost混合模型，实现番茄叶片病害精准检测与分类。","YOLOv8定位病斑，Inception-v4提取多尺度特征，XGBoost降维","10类病害识别精度超现有方法，训练准确率0.9969，测试0.9695，单图CPU约200毫秒。","可探索IoT边缘部署、多模态数据融合与可解释AI，提升田间实时病害诊断可信度。","智慧农业 \u002F 农业物联网","2026-09-10T23:30:14.735468Z",{"id":406,"title":407,"url":408,"summary":409,"summary_zh":410,"content":9,"source_name":411,"source_url":408,"published_at":412,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":413,"score_detail":414,"sources":416,"tags":418,"primary_topic":421,"secondary_topics":422,"classification":424,"search_phrases":440,"slug":443,"view_count":78,"doi":444,"paper":445,"created_at":459},5387,"LIGHTWEIGHT FEATURE FUSED ATTENTION DRIVEN MULTI LEVEL HYBRID MOBILENET V2 FRAMEWORK FOR EARLY TOMATO LEAF DISEASE DETECTION","https:\u002F\u002Fdoi.org\u002F10.62441\u002Factainnovations.vi.1918","It is crucial to find and treat tomato crop diseases early and accurately to improve crop efficiency and make sure there is no scarcity of crop. However, computerized analysis is provocative due to lighting differences, compound backgrounds, and graphic likenesses among infection regions. Research work proposed a lightweight Attention Driven Multi Level Hybrid MobileNet-V2 algorithm that fuses the handcrafted texture structures with Deep Learning (DL) features, then image preprocessing is incorporated using Gaussian, Bilateral, and Guided filters to minimize noise by preserving the disease lesions. LAB color model along with Contrast Limited Adaptive Histogram Equalization (CLAHE) are used to improve the contrast. Otsu thresholding is involved to effectively separate the leaf segment from diseased parts. Local Binary Pattern (LBP) and Gray Level Co-occurrence Matrix (GLCM) descriptors are used to obtain fine-grained features, while high level features are gained by pretrained MobileNet-V2 model. Features fusion is fed to the fully connected layers of deep learning model for training and validation purpose. Results have revealed that the suggested model attains better classification performance as compared to existing models, making it an effective explanation for real time tomato leaf disease detection for smart agriculture applications.","早期准确发现并治疗番茄作物病害，对于提高作物效率、确保作物不短缺至关重要。然而，由于光照差异、复杂背景以及感染区域之间的图形相似性，计算机化分析具有挑战性。研究提出了一种轻量级注意力驱动多层混合MobileNet-V2算法，该算法将手工纹理结构与深度学习（DL）特征相融合，然后结合高斯、双边和引导滤波器进行图像预处理，以在保留病害病变的同时最小化噪声。采用LAB颜色模型和对比度受限自适应直方图均衡化（CLAHE）来提高对比度。引入Otsu阈值分割以有效分离叶片片段与病害部分。使用局部二值模式（LBP）和灰度共生矩阵（GLCM）描述符获取细粒度特征，同时通过预训练的MobileNet-V2模型获取高层特征。特征融合后输入深度学习模型的全连接层进行训练和验证。结果表明，与现有模型相比，所提出的模型取得了更好的分类性能，使其成为智慧农业应用中实时番茄叶片病害检测的有效解决方案。","Acta Innovations","2026-10-18T00:00:00Z",64,{"impact":113,"substance":99,"depth":19,"authority":13,"freshness":17,"relevant":20,"comment":415},"提出融合手工纹理特征与MobileNet-V2的轻量番茄叶病检测方法，有一定方法新颖性，但属单篇论文、应用范围有限，未达每日精选门槛。",[417],{"name":411,"url":408},[25,26,302,419,420],"番茄病害","图像识别",{"code":31,"slug":32,"name":33,"path":34,"parent_code":35},[423],{"code":60,"slug":223,"name":224,"path":225,"parent_code":226},{"status":38,"locked":14,"version":39,"method":40,"evidence":425},{"candidates":426,"agricultural_evidence":45},[427,431,435],{"code":31,"score":44,"supported":45,"title":428,"abstract":429,"tags":430},[47],[49,47],[419],{"code":60,"score":161,"supported":45,"title":432,"abstract":433,"tags":434},[],[302,318],[26,420,302],{"code":52,"score":53,"supported":45,"title":436,"abstract":437,"tags":439},[],[25,438],"smart agriculture",[25],[441,442],"MobileNet-V2 番茄叶片病害","番茄叶病 深度学习 检测","MobileNet-V2番茄叶片病害-5387","10.62441\u002Factainnovations.vi.1918",{"doi":444,"openalex_id":446,"authors":447,"venue":411,"cited_by_count":78,"oa_url":453,"card":454,"direction":403,"ingested_from":85},"W7222977034",[448,450],{"name":449,"orcid":9},"Mrs. Vijetalaxmi A Hegde",{"name":451,"orcid":452},"Abhay Ramakant Kasetwar","https:\u002F\u002Forcid.org\u002F0000-0002-5800-8764","https:\u002F\u002Factainnovations.com\u002Findex.php\u002Fpub\u002Farticle\u002Fdownload\u002F1918\u002F981",{"tldr":455,"method":456,"finding":457,"direction":83,"opportunity":458},"提出轻量注意力多级混合MobileNet-V2框架，融合手工纹理与深度学习特征实现番茄叶片病害早期检","高斯\u002F双边\u002F引导滤波去噪，LAB+CLAHE增强，Otsu分割，LBP与GLCM","所提模型分类性能优于现有模型，可有效用于智慧农业实时番茄叶片病害检测。","可探索多作物跨域泛化、田间复杂光照下轻量模型部署与可解释性，及病害早期分级检测。","2026-10-20T23:30:10.424511Z"]