TripleS: Mitigating multi-task learning conflicts for semantic change detection in high-resolution remote sensing imagery

About These are the constructed datasets in CVEO's recent research paper published on ISPRS Journal of Photogrammetry and Remote Sensing. This repository contains two full-coverage SCD datasets, namely SC-SCD7 and CC-SCD5. The SC-SCD7 dataset includes seven land-cover categories: bareland, water, building, structure, farmland, vegetation, and road. The CC-SCD5 dataset encompasses five land-cover categories: water, building, road, vegetation and others, the latter category including areas such as bareland, cropland, and other multiplicate landforms. For more details, please refer to our paper and visit our GitHub repo. Citation Please consider citing the following paper if you used this project and dataset in your research: @article{TAN2025374, title = {TripleS: Mitigating multi-task learning conflicts for semantic change detection in high-resolution remote sensing imagery}, journal = {ISPRS Journal of Photogrammetry and Remote Sensing}, volume = {230}, pages = {374-401}, year = {2025}, issn = {0924-2716}, doi = {https://doi.org/10.1016/j.isprsjprs.2025.09.019}, url = {https://www.sciencedirect.com/science/article/pii/S0924271625003776}, author = {Xiaoliang Tan and Guanzhou Chen and Xiaodong Zhang and Tong Wang and Jiaqi Wang and Kui Wang and Tingxuan Miao}, keywords = {Semantic change detection, Remote sensing, Multi-task learning, Deep learning, Land-cover and land-use}}