SCanNet

Pytorch codes of Joint Spatio-Temporal Modeling for Semantic Change Detection in Remote Sensing Images [paper]

Checkpoints

For readers to easily evaluate the accuracy, we provide the trained weights.

SECOND:

1.Drive
2.Baidu (pswd: SCAN)

LandsatSCD:

1.Drive
2.Baidu (pswd: SCAN)

Landsat-SCD

The land-scd dataset needs to be pre-processed to meet the experimental settings in this paper. More details are provided at /datasets/LandsatSCD/read_me.md

For readers' convenience, we also provide the preprocessed data:

Baidu Netdisk (psswd lscd)

Google Drive

Cite SCanNet

If you find this work useful or interesting, please consider citing the following BibTeX entry.

@article{ding2024joint,
  title={Joint Spatio-Temporal Modeling for Semantic Change Detection in Remote Sensing Images},
  author={Ding, Lei and Zhang, Jing and Guo, Haitao and Zhang, Kai and Liu, Bing and Bruzzone, Lorenzo},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  year={2024},
  volume={62},
  pages={1-14},
  doi={10.1109/TGRS.2024.3362795}
}

(Note: This repository is under construction, contents are not final.)