Home

Awesome

ASDDPM-Adaptive-Semantic-Enhanced-DDPM

The code for Adaptive Semantic-Enhanced Denoising Diffusion Probabilistic Model for Remote Sensing Image Super-Resolution, the other models shown in the experiments could be refered to Remote-Sensing-Super-resolution-Model-Collection.

Models

Dataset

The experimental datasets, OLI2MSI and Alsat, could be obtained from:

Usage

Train

  1. Download the RRDB pre-train model from link and put it into the right place (./models/LREncoder/) according to the guidance in this link. https://pan.baidu.com/s/1siWepPn2pVFC3SGyk1wuJg code：bean
  2. Change the model name and data information in the option.py

python src/main.py

Test

  1. Put pre-trained model into 'pre_train'
  2. Change the model name in the option.py

python test.py

Weight

Our pre-train ASDDPM and RRDB model on OLI2MSI and ALSAT could be downloaded from link：https://pan.baidu.com/s/1siWepPn2pVFC3SGyk1wuJg code：bean

A guidance file is also shared in this link, please put the pre-train model to the right place according to the guidance. (Attention: the code could run only after the RRDB pre-train model is put in the right place.)

Cite

@article{sui2024adaptive,
  title={Adaptive Semantic-Enhanced Denoising Diffusion Probabilistic Model for Remote Sensing Image Super-Resolution},
  author={Sui, Jialu and Ma, Xianping and Zhang, Xiaokang and Pun, Man-On},
  journal={arXiv preprint arXiv:2403.11078},
  year={2024}
}
@article{sui2023gcrdn,
  title={Gcrdn: Global context-driven residual dense network for remote sensing image super-resolution},
  author={Sui, Jialu and Ma, Xianping and Zhang, Xiaokang and Pun, Man-On},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  year={2023},
  publisher={IEEE}
}
@article{sui2024denoising,
  title={Denoising Diffusion Probabilistic Model with Adversarial Learning for Remote Sensing Super-Resolution},
  author={Sui, Jialu and Wu, Qianqian and Pun, Man-On},
  journal={Remote Sensing},
  volume={16},
  number={7},
  pages={1219},
  year={2024},
  publisher={MDPI}
}

Be free to create a issue if you have any questions! I will reply as soon as possible!