PUMCL

This is a repository for releasing a PyTorch implementation of our work Prototypical Unknown-Aware Multiview Consistency Learning for Open-Set Cross-Domain Remote Sensing Image Classification.

If you find this repository useful for your research, please cite our paper:

@ARTICLE{10707611,
  author={Zhang, Xiaokang and Wu, Wanjing and Zhang, Mi and Yu, Weikang and Ghamisi, Pedram},
  journal={IEEE Transactions on Geoscience and Remote Sensing}, 
  title={Prototypical Unknown-Aware Multiview Consistency Learning for Open-Set Cross-Domain Remote Sensing Image Classification}, 
  year={2024},
  volume={62},
  number={},
  pages={1-16},
  keywords={Remote sensing;Training;Adaptation models;Entropy;Sensors;Uncertainty;Scene classification;Generators;Adversarial machine learning;Noise measurement;Consistency learning;open-set;remote sensing images;scene classification;unsupervised domain adaptation (UDA)},
  doi={10.1109/TGRS.2024.3476151}}

Acknowledgements

This codebase is heavily borrowed from SSOUDA and MAOSDAN.