[论文] The Last Remote Sensing Model Built by Humans? Casting Any Earth Obser...

论文概要

研究领域: NLP
作者: Sijie Zhao, Zeqi Zhou, Heli Qi, Xuanchao Fu, Zhuo Zheng, Junshi Xia, Junjue Wang, Weihao Xuan, Naoto Yokoya
发布时间: 2026-10-13
arXiv: 2610.18845

中文摘要

地球观测（EO）模型仍由人工构建。对于每个新任务，专家设计训练方案并对照基准进行调优。我们证明这个循环可以交给 agent。任何可验证任务都可以表示为 harness（脚手架）和验证器，harness 定义 agent 可用的资源，验证器定义成功的标准。每个 EO 任务都可以通过带标注基准、独立传感器、物理模型或对同一地点的重复观测来变得可验证。仅给定这两个组件，通用编码 agent 即可构建 EO 模型并在自身输出上递归改进。长期被视为独立研究问题的设置——从少量标签学习到跨域自适应——变成了 harness 和验证器的选择问题。在涵盖光学、SAR 和多光谱时序图像的四个 EO 基准上，agent 构建的模型超越了现有最优，且仅用约 1% 的训练标签就保持了大部分精度。这确立了 EO 的新范式：人设计 harness 和验证器，agent 构建模型，EO 模型随 agent 一起进步。

原文摘要

Earth observation (EO) models are still built by hand. For each new任务， experts design a training recipe and tune it against a benchmark. We show that this loop can be handed to an agent. Any verifiable task can be cast as a harness and a verifier, where the harness defines what the agent may use and the verifier defines what counts as success. Every EO task can be made verifiable, through labeled benchmarks, independent sensors, physical models or the repeated observation of the same place. Given only these two components, a general-purpose coding agent builds the EO model and improves it recursively on its own outputs. Settings long treated as separate research problems, from learning with a handful of labels to adapting across domains, become choices of harness and verifier. On four EO b...

自动采集于 2026-10-15

#论文 #arXiv #NLP #小凯