发表机构
Laboratoire des Sciences du Climat et de l’Environnement（气候与环境科学实验室）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对中等分辨率卫星影像的冠层高度预测，探索像素级注意力机制，发现其优于大补丁方法，并通过高效注意力变体平衡质量与资源，为像素级回归任务提供设计指导。
AI 中文摘要
从中等分辨率卫星影像预测冠层高度是评估全球森林状况的一种常见且可扩展的方法，而森林在减缓气候变化中发挥着至关重要的作用。尽管基于Transformer的架构在许多领域表现出强大的性能，但将其直接应用于密集（即像素级）回归任务往往会产生次优结果。特别是，补丁大小对模型性能有着至关重要的影响。在这项工作中，我们考虑了像素级注意力机制，并表明由此产生的模型通常优于依赖较大补丁大小的模型。然而，像素级注意力可能是一种资源消耗过大的操作。为此，我们使用高效的注意力变体进行了广泛的实验研究，以确定预测质量与资源需求之间的有利权衡，从而促进所提出模型的实际部署。此外，我们与该领域几种成熟的模型进行了全面比较，并表明在合适的超参数选择下，基于Transformer的架构可以超越竞争方法。我们的研究结果为中等分辨率卫星影像上的像素级回归任务（包括冠层高度和生物量估算、土壤水分制图和产量预测）设计模型提供了实用指导。
英文摘要
Predicting canopy height from medium-resolution satellite imagery is a common and scalable approach for assessing the condition of the world's forests, which play a crucial role in climate change mitigation. While Transformer-based architectures have shown strong performance in many domains, their straightforward application to dense (i.e., pixel-level) regression tasks often yields suboptimal results. In particular, the patch size has a crucial impact on the model performance. In this work, we consider pixel-level attention schemes and show that the resulting models generally outperform those relying on larger patch sizes. However, pixel-level attention can be a prohibitively resource-intensive operation. For this reason, we conduct an extensive experimental study using efficient attention variants to identify favorable trade-offs between prediction quality and resource requirements, facilitating the practical deployment of the proposed models. In addition, we perform a comprehensive comparison with several well-established models in the field and show that, with suitable hyperparameter choices, Transformer-based architectures can outperform competing approaches. Our findings provide practical guidance for designing models for pixel-level regression tasks on medium-resolution satellite imagery, including canopy height and biomass estimation, soil moisture mapping, and yield forecasting.
CommentsAccepted at ACM SIGSPATIAL 2026
