Dataset

The dataset comprises data from two cities, intended for evaluating the pretraining and transfer-learning capabilities of the GitHub repository stark-t/UncertaintySlumDetection (github.com). Each city's dataset consists of three key components: the Sentinel-2 RGB image, resampled to a 3m ground sample distance (GSD), along with urban and background masks, and the slum reference mask

本数据集涵盖两座城市的相关数据，旨在评估GitHub仓库stark-t/UncertaintySlumDetection（github.com）的预训练与迁移学习能力。各城市的数据集均包含三个核心组成部分：重采样至3米地面采样距离（Ground Sample Distance，GSD）的Sentinel-2 RGB影像、城市与背景掩膜，以及贫民窟参考掩膜。