Meta-Learning for Data-Efficient Plant Growth Estimation via Vision Transformers and Fuzzy Clustering
本文通过结合视觉变换器特征嵌入、基于模糊聚类的任务构建和基于梯度的元学习,提出了一种少样本回归框架以解决植物生长估计中标签数据稀缺的问题。
本文通过结合视觉变换器特征嵌入、基于模糊聚类的任务构建和基于梯度的元学习,提出了一种少样本回归框架以解决植物生长估计中标签数据稀缺的问题。
为解决在线测试时适应中的漂移或崩溃问题,提出SEGA方法,通过敏感性引导擦除适应,提高模型在分布偏移下的鲁棒性和稳定性。
为解决开放词汇语义分割中持续测试时分布偏移导致的视觉-语言对齐脆弱问题,提出DAF框架,通过增加边缘多样性损失、跨模态锚点一致性损失和特征显著性过滤来稳定模型。
This study addresses the absence of target samples in zero-shot learning by proposing a training-free, image-free analytical semantic transfer framework. Through a closed-form weight injection mechanism, pretrained classifiers are extended to unseen categories without iterative optimization. Furthermore, this work establishes a finite-sample error decomposition theory and introduces a computable semantic extrapolation residual metric to guide data construction. Empirical evaluations on standard benchmarks demonstrate that the proposed method matches or surpasses existing image-free approaches while approaching few-shot performance levels. Collectively, this framework provides an efficient analytical solution with rigorous theoretical support for zero-shot learning, effectively bridging the gap between training-free efficiency and high-accuracy generalization in the absence of visual data.
This study addresses the challenge of estimating optimal doses in nonlinear dose–response relationships, particularly in contexts such as aquaculture, by introducing a Bayesian fractional polynomial framework for the first time. The proposed approach explicitly quantifies and integrates model uncertainty through Bayesian model averaging, thereby enhancing the robustness and accuracy of optimal dose estimation. In simulation studies, the method significantly outperforms existing benchmark approaches. Furthermore, when applied to real-world data on fish nutritional requirements, it successfully identifies the optimal nutrient dose level, demonstrating both practical applicability and statistical reliability.
本文通过结合视觉变换器特征嵌入、基于模糊聚类的任务构建和基于梯度的元学习,提出了一种少样本回归框架以解决植物生长估计中标签数据稀缺的问题。
为解决在线测试时适应中的漂移或崩溃问题,提出SEGA方法,通过敏感性引导擦除适应,提高模型在分布偏移下的鲁棒性和稳定性。
为解决开放词汇语义分割中持续测试时分布偏移导致的视觉-语言对齐脆弱问题,提出DAF框架,通过增加边缘多样性损失、跨模态锚点一致性损失和特征显著性过滤来稳定模型。
This study addresses the absence of target samples in zero-shot learning by proposing a training-free, image-free analytical semantic transfer framework. Through a closed-form weight injection mechanism, pretrained classifiers are extended to unseen categories without iterative optimization. Furthermore, this work establishes a finite-sample error decomposition theory and introduces a computable semantic extrapolation residual metric to guide data construction. Empirical evaluations on standard benchmarks demonstrate that the proposed method matches or surpasses existing image-free approaches while approaching few-shot performance levels. Collectively, this framework provides an efficient analytical solution with rigorous theoretical support for zero-shot learning, effectively bridging the gap between training-free efficiency and high-accuracy generalization in the absence of visual data.
This study addresses the challenge of estimating optimal doses in nonlinear dose–response relationships, particularly in contexts such as aquaculture, by introducing a Bayesian fractional polynomial framework for the first time. The proposed approach explicitly quantifies and integrates model uncertainty through Bayesian model averaging, thereby enhancing the robustness and accuracy of optimal dose estimation. In simulation studies, the method significantly outperforms existing benchmark approaches. Furthermore, when applied to real-world data on fish nutritional requirements, it successfully identifies the optimal nutrient dose level, demonstrating both practical applicability and statistical reliability.