OlmoEarth Studio 推出自定义嵌入导出功能
原标题:Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis 1 August 12, 2026
AI 摘要
AI2 的 OlmoEarth Studio 平台推出自定义嵌入向量导出功能,用户可计算并下载地球观测数据的紧凑数值表示。这些嵌入向量支持相似性搜索、分割和无监督探索等下游任务,并可通过 Studio UI 或 API 定制区域、时间、编码器变体等参数。官方展示了基于 60 个标注像素的线性分类器在越南红树林区域达到 0.84 的加权 F1 分数,证明了嵌入向量的有效性。
正文节选
Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis 📄 Tech Report: https://allenai.org/papers/olmoearth | 📊 Documentation: https://docs.olmoearth.allenai.org/embeddings | 💻 Learn more about OlmoEarth: https://allenai.org/olmoearth OlmoEarth Studio, our platform for building Earth observation models, now lets you compute and export embedding vectors—compact numerical representations of Earth-observation data produced by our open source OlmoEart