GVLM 是全球首个大规模、开源的 VHR（超高分辨率）滑坡测绘数据集，它不仅为滑坡研究提供了宝贵的数据资源，更为变化检测算法在复杂自然场景下的鲁棒性测试提供了新的基准。覆盖面积 163.77km²，空间分辨率0.59m，每个子数据集均包含一对双时态图像及相应的地面实况地图。", "ds_source": "
数据来源于Google Earth。", "ds_process_way": "
依托多目标域自适应技术，配套原型引导域感知渐进表征学习（PG-DPRL）方法，无需额外样本标注即可构建适用于多目标域的通用滑坡制图模型。数据集采用由近及远自适应策略，兼顾域间特征差异，结合跨域原型学习生成可靠域专属伪标签，聚合跨域特征习得共享决策边界；引入 Wasserstein 距离度量与跨域原型一致性损失，实现类别级特征对齐并提升表征判别性。", "ds_quality": "
数据集包含全球 17 处不同滑坡场地异构子域的超高分辨率遥感数据（GVLM 数据集），实验验证该数据集适配模型迁移性能优异，在全目标域总体精度与平均精度指标上优于主流前沿算法，可为大区域、跨场景无标注遥感滑坡自动制图及灾害风险防控提供可靠数据支撑。", "ds_acq_start_time": "2010-03-20 00:00:00", "ds_acq_end_time": "2021-04-15 00:00:00", "ds_acq_place": "全球", "ds_acq_lon_east": null, "ds_acq_lat_south": null, "ds_acq_lon_west": null, "ds_acq_lat_north": null, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "open-access", "ds_total_size": 1049200128, "ds_files_count": 2, "ds_format": "png", "ds_space_res": "0.59m", "ds_time_res": "", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "e274471b-79d2-4e9e-95da-f79d6c214e7d.jpg", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "52b7b79b-860c-49a5-9083-9a70cf8bed5a", "ds_serv_man": "李红星", "ds_serv_phone": "0931-4967592", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.45", "170.50" ], "quality_level": 3, "publish_time": "2026-03-25 10:02:55", "last_updated": "2026-03-25 10:33:25", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.NCDC.LANDSLIDE.DB7211.2026", "i18n": { "en": { "title": "Global Very-High-Resolution Landslide Mapping（GVLM） Dataset Version 1.0", "ds_format": "", "ds_source": "
=The data is sourced from Google Earth.", "ds_quality": "", "ds_ref_way": "", "ds_abstract": "
GVLM is the world's first large-scale, open-source VHR (ultra-high resolution) landslide mapping dataset, which not only provides valuable data resources for landslide research, but also provides a new benchmark for robust testing of change detection algorithms in complex natural scenes. The coverage area is 163.77km ², with a spatial resolution of 0.59m. Each sub dataset contains a pair of dual temporal images and corresponding ground truth maps.", "ds_time_res": "", "ds_acq_place": "", "ds_space_res": "", "ds_projection": "", "ds_process_way": "", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "license_type": "https://creativecommons.org/licenses/by/4.0/", "doi_reg_from": "reg_local", "cstr_reg_from": "reg_local", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "ds_topic_tags": [ "滑坡", "高分辨率", "全球" ], "ds_subject_tags": [ "地理学", "地质学" ], "ds_class_tags": [], "ds_locus_tags": [ "全球" ], "ds_time_tags": [ 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021 ], "ds_contributors": [ { "true_name": "张效康", "email": "zhangxiaokang@whu.edu.cn", "work_for": "武汉大学", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "张效康", "email": "zhangxiaokang@whu.edu.cn", "work_for": "武汉大学", "country": "中国" } ], "ds_managers": [ { "true_name": "张效康", "email": "zhangxiaokang@whu.edu.cn", "work_for": "武汉大学", "country": "中国" } ], "category": "灾害" }