Satellite Image Semantic Segmentation
This addresses land cover classification for remote sensing applications, but it is incremental as it applies an existing method to new data.
The paper tackles semantic segmentation of satellite images into six land cover classes using a Swin Transformer architecture on a dataset built from IGN open data, reporting quantitative and qualitative results and making the dataset and model publicly available.
In this paper, we propose a method for the automatic semantic segmentation of satellite images into six classes (sparse forest, dense forest, moor, herbaceous formation, building, and road). We rely on Swin Transformer architecture and build the dataset from IGN open data. We report quantitative and qualitative segmentation results on this dataset and discuss strengths and limitations. The dataset and the trained model are made publicly available.