Citation
If you find it useful, please consider citing:
@article{wang2026rsgpnet,
title={RSGPNet: Geometric Prompting for Remote Sensing Open-Vocabulary Semantic Segmentation},
author={Wang, Shanwen and Sun, Xin and Wang, Sirui and Zhu, Xiao Xiang},
journal={arXiv preprint arXiv:2606.28410},
year={2026}
}
Acknowledgments
We sincerely thank the authors of SAM3 and SegEarth‑OV3 for their excellent open‑source work, and we also thank the contributors of the publicly available LoveDA, Potsdam, UDD5 and Vaihingen datasets. Please follow the licenses and terms of the original models and datasets.