发表机构
Technical University of Denmark (DTU); Pioneer Centre for Artificial Intelligence（丹麦技术大学（DTU）; 人工智能先锋中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对跨物种三维植物器官分割中固定空间粒度泛化能力差的问题，提出结合冻结Utonia基础模型与自适应粒度选择的小样本方法AGS-PlantSeg，在三个数据集上实现88.9%平均mIoU，性能优于固定粒度基线且与全监督架构相当。
AI 中文摘要
近期的三维基础模型通过控制空间粒度，为点云学习提供了强大的特征表示。然而，依赖固定空间粒度严重限制了植物表型分析等应用中的泛化能力，在这类应用中，器官的形态和大小会随物种及生长阶段发生显著变化。为解决这一问题，我们提出了AGS-PlantSeg，这是一种小样本三维植物器官分割方法，它利用冻结的Utonia（arXiv:2603.03283）基础模型，并结合自适应粒度选择技术。通过为每种特定植物模型动态选择最佳粒度级别，我们的方法为轻量级MLP分割头提取了优化的几何特征。在PLANesT-3D（arXiv:2407.21150）、Pheno4D和Crops3D数据集上进行的大量实验表明，AGS-PlantSeg显著提升了跨物种泛化能力，达到了88.9%的平均mIoU性能，且比固定粒度基线高出2.5个mIoU点。尽管仅需要极少的标注数据，我们的方法与全监督的植物专用架构相比仍极具竞争力。
英文摘要
Recent 3D foundation models provide powerful feature representations for point cloud learning by controlling spatial granularity. However, relying on a fixed spatial granularity severely limits generalization in applications like plant phenotyping, where organ morphology and size vary substantially across species and growth stages. To address this, we propose AGS-PlantSeg, a few-shot 3D plant organ segmentation method that leverages the frozen Utonia (arXiv:2603.03283) foundation model combined with Adaptive Granularity Selection. By dynamically selecting the best granularity levels for each specific plant model, our method extracts optimized geometric features for a lightweight MLP segmentation head. Extensive experiments across PLANesT-3D (arXiv:2407.21150), Pheno4D , and Crops3D demonstrate that AGS-PlantSeg significantly improves cross-species generalization, achieving 88.9% average mIoU performance and outperforming fixed-granularity baselines by 2.5 mIoU points. Despite requiring minimal annotated data, our approach is highly competitive with fully supervised, plant-specific architectures.
CommentsAccepted at the Computer Vision in Plant Phenotyping and Agriculture (CVPPA) Workshop at the European Conference on Computer Vision (ECCV) 2026. Project page: https://dtu-pas.github.io/ags-plantseg/
