标题：Mapping the World at Very High Resolution: Toward a New Era of Global Land Intelligence
报告时间：2026年6月15日（星期一）15:30-17:00
报告地点：人民大街校区地理科学学院322会议室
主讲人：夏俊士
主办单位：地理科学学院
报告内容简介：
Starting from OpenEarthMap and its extension OpenEarthMap-SAR, this talk presents recent advances in geospatial AI, foundation models, and multimodal Earth observation. We introduce benchmark datasets including DisasterM3, DynamicVL, BRIGHT, and SynRS3D, highlighting their roles in high-resolution land cover mapping, 3D city understanding, disaster damage assessment, and dynamic Earth monitoring. These developments demonstrate how next-generation geospatial foundation models are enabling scalable, accurate, and actionable global land intelligence for a wide range of environmental and societal applications.
主讲人简介：
夏俊士，现任日本理化学研究所终身高级研究员，长期致力于遥感人工智能与全球地理空间智能研究。聚焦于高分辨率土地覆盖制图、多模态遥感数据融合、地理空间基础模型、灾害智能感知与大规模环境监测等方向，旨在构建面向全球尺度的智能地表认知与制图技术体系。其研究成果推动了高分辨率土地覆盖数据集、跨模态遥感基准和新一代地理空间智能模型的发展。目前，担任 IEEE Geoscience and Remote Sensing Letters、Remote Sensing等期刊副主编，同时担任 IEEE Signal Processing Magazine 高级编委。
