AllClear 公共云层去除数据集
收藏资源简介:
卫星图像中的云层对于下游应用构成了重大挑战,当前云去除研究面临的一个主要问题是缺乏全面的基准测试和足够大且多样化的训练数据集。为了解决这一问题,康奈尔大学和哥伦比亚大学的研究团队于 2024 年推出了 AllClear,这是目前最大的公共云层去除数据集,包含 23,742 个全球分布的兴趣区域 (ROIs),覆盖了多样的土地利用模式,总共包含 400 万张图像。相关论文成果为「AllClear: A Comprehensive Dataset and Benchmark for Cloud Removal in Satellite Imagery」,已被 NeurIPS 接受。
Cloud cover in satellite imagery poses significant challenges for downstream applications. A major hurdle in current cloud removal research is the lack of comprehensive benchmarks and sufficiently large, diverse training datasets. To address this issue, a research team from Cornell University and Columbia University launched AllClear in 2024, which is currently the largest public cloud removal dataset. It contains 23,742 globally distributed regions of interest (ROIs) covering diverse land use patterns, with a total of 4 million images. The associated paper titled "AllClear: A Comprehensive Dataset and Benchmark for Cloud Removal in Satellite Imagery" has been accepted by NeurIPS.




