发表机构
Aerospace Information Research Institute, Chinese Academy of Sciences; University of Chinese Academy of Sciences; The Hong Kong University of Science and Technology (Guangzhou); Helmholtz-Zentrum Dresden-Rossendorf（中国科学院空天信息创新研究院; 中国科学院大学; 香港科技大学（广州）; 德累斯顿-罗森多夫亥姆霍兹中心）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出SimCRAFT框架，通过合成轨迹与上下文检索增强微调，将复杂遥感编排能力蒸馏到7B模型，性能优于开源LLM，为轻量遥感智能提供开源基线。
AI 中文摘要
地球观测数据的规模与多样性前所未有的激增，暴露了传统人工工作流程的关键瓶颈，推动了遥感（RS）智能体的出现。然而，这些先进智能体的实际部署因严重依赖大型通用大语言模型（LLM）而受阻，这类模型缺乏深厚的领域专业知识，且基础设施要求高得难以承受。为解决该问题，我们提出SimCRAFT，这是一个与模型无关的框架，可将复杂的RS编排能力蒸馏到7B规模的紧凑模型中。针对数据稀缺问题，我们首先将多智能体合成引擎与模拟执行引擎配对，后者会检查模式正确性、工具间依赖关系以及传感器/工具兼容性，生成SimRS-14k——一个大规模、经约束验证的工作流规划语料库。其次，我们提出上下文检索增强微调（CRAFT），该方法通过在抗噪目标下将检索到的标准操作程序适配到新查询，对模型进行微调以实现类比推理，将检索增强微调（RAFT）推广到多步骤RS工作流规划，且不会出现机械复制问题。大量实验表明，SimCRAFT-7B显著优于开源LLM，可与先进的闭源模型及专用RS智能体相媲美，且在三种7B骨干网络上均可复现。本研究为轻量RS智能提供了具有竞争力的开源基线，支持在资源受限或资源节约条件下实现高效自主部署。
英文摘要
The unprecedented surge in Earth observation data volume and diversity has exposed a critical bottleneck for traditional manual workflows, catalyzing the emergence of Remote Sensing (RS) Agents. However, the practical deployment of these advanced agents is severely hindered by their heavy reliance on large-scale general-purpose LLMs, which lack deep domain expertise and impose prohibitive infrastructure demands. To resolve this, we propose SimCRAFT, a model-agnostic framework that distills sophisticated RS orchestration capabilities into a compact 7B-scale model. Addressing data scarcity, we first pair a multiagent synthesis engine with a Mock Execution Engine that checks schema correctness, inter-tool dependencies, and sensor/tool compatibility, producing SimRS-14k, a large-scale, constraint-validated workflow planning corpus. Second, we propose Contextual Retrieval-Augmented Fine-Tuning (CRAFT) that finetunes the model to reason analogically by adapting retrieved Standard Operating Procedures to novel queries under a noise-robust objective, generalizing RAFT to multi-step RS workflow planning without mechanical copying. Extensive experiments demonstrate that SimCRAFT-7B significantly outperforms openweights LLMs and rivals advanced closedsource models and specialized RS agents, while reproducing across three 7B backbones. This work contributes a competitive open-weights baseline for lightweight RS intelligence, enabling efficient autonomous deployment under resource-constrained or resource-conserving conditions.
CommentsAccepted by EMNLP 2026 as a Main Conference paper
