[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2316":3},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":23,"tags":25,"view_count":31,"doi":32,"paper":33,"created_at":60},2316,"Vision Transformers Enable Advanced Plant Phenotyping in Controlled Environments","https:\u002F\u002Fdoi.org\u002F10.64898\u002F2026.09.04.748299","Reliable plant segmentation in high-throughput phenotyping must transfer across species and imaging conditions without repeated model tuning or extensive reannotation. We compare three segmentation strategies using images from Oak Ridge National Laboratory's Advanced Plant Phenotyping Laboratory: (i) fixed color-based thresholding, (ii) supervised U-Nets trained from scratch, and (iii) pretrained vision transformers fine-tuned for binary segmentation. Models were evaluated on a held-out test set and a generalization set that comprised unseen species. On the held-out test set, thresholding, the best U-Net, and the best vision transformer achieved mean Dice scores of 58.3, 96.6, and 97.3, respectively. On the generalization set, the corresponding Dice scores were 56.5, 86.2, and 95.7. Thresholding remained effective on some datasets but failed when plant appearance changed. Supervised U-Net training resolved within-distribution errors but failed to generalize to novel species and backgrounds. Pretrained vision transformers consistently produced high-accuracy segmentations across the evaluated species, views, soil backgrounds, and tray types. These results benchmark the practical progression from fixed rules to task-specific supervision and pretrained visual representations for controlled-environment plant phenotyping.","在高通量表型分析中，可靠的植物分割必须能够在不同物种和成像条件之间迁移，而无需重复调整模型或进行大量重新标注。我们利用橡树岭国家实验室先进植物表型实验室的图像，比较了三种分割策略：（i）基于颜色的固定阈值法，（ii）从零开始训练的有监督U-Net，以及（iii）针对二值分割进行微调的预训练视觉Transformer。模型在留出测试集和包含未见物种的泛化集上进行了评估。在留出测试集上，阈值法、最佳U-Net和最佳视觉Transformer的平均Dice分数分别为58.3、96.6和97.3。在泛化集上，相应的Dice分数分别为56.5、86.2和95.7。阈值法在某些数据集上仍然有效，但当植物外观发生变化时则失效。有监督U-Net训练解决了分布内误差，但未能泛化到新物种和背景。预训练视觉Transformer在所评估的物种、视角、土壤背景和托盘类型上始终产生高精度分割。这些结果基准了受控环境植物表型分析中从固定规则到任务特定监督再到预训练视觉表示的实际进展。",null,"bioRxiv (Cold Spring Harbor Laboratory)","2026-09-10T00:00:00Z","论文",10,false,79,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,22,13,8,1,"预训练视觉Transformer在跨物种植物分割上显著优于U-Net与阈值法，为受控环境高通量表型提供可复用基准，方法新颖、数据扎实，值得入选。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","农业人工智能","高通量表型","植物表型","图像分割",0,"10.64898\u002F2026.09.04.748299",{"doi":32,"openalex_id":34,"authors":35,"venue":10,"cited_by_count":31,"oa_url":9,"card":53,"direction":57,"ingested_from":59},"W7212309344",[36,38,41,44,47,50],{"name":37,"orcid":9},"Janou Milligan",{"name":39,"orcid":40},"Anand Seethepalli","https:\u002F\u002Forcid.org\u002F0000-0003-0937-9128",{"name":42,"orcid":43},"Aristeidis Tsaris","https:\u002F\u002Forcid.org\u002F0000-0002-7734-3349",{"name":45,"orcid":46},"Xiao Wang","https:\u002F\u002Forcid.org\u002F0000-0001-6545-1943",{"name":48,"orcid":49},"Larry M. York","https:\u002F\u002Forcid.org\u002F0000-0002-1995-9479",{"name":51,"orcid":52},"John Lagergren","https:\u002F\u002Forcid.org\u002F0000-0002-8092-7433",{"tldr":54,"method":55,"finding":56,"direction":57,"opportunity":58},"比较阈值法、U-Net和预训练ViT在植物分割中的跨物种泛化能力。","使用ORNL表型实验室图像，对比阈值法、U-Net和微调ViT的分割性能。","预训练ViT在未见物种上Dice达95.7，显著优于U-Net的86.2和阈值法的56.5。","农业遥感与作物表型","可探索预训练ViT在田间复杂场景的跨物种泛化，并降低微调数据需求。","openalex","2026-09-13T23:30:21.474320Z"]