https://github.com/xavierjiezou/pmaa

Official PyTorch implementation of "PMAA: A Progressive Multi-scale Attention Autoencoder Model for High-Performance Cloud Removal from Multi-temporal Satellite Imagery" (ECAI 2023).
https://github.com/xavierjiezou/pmaa

attention autoencoder cloud-removal generative-model image-processing image-restoration lightweight low-level-vision remote-sensing

Last synced: about 2 months ago
JSON representation

Official PyTorch implementation of "PMAA: A Progressive Multi-scale Attention Autoencoder Model for High-Performance Cloud Removal from Multi-temporal Satellite Imagery" (ECAI 2023).

• Host: GitHub
• URL: https://github.com/xavierjiezou/pmaa
• Owner: XavierJiezou
• License: mit
• Created: 2023-07-30T04:17:29.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-08-23T01:56:19.000Z (over 1 year ago)
• Last Synced: 2024-01-29T20:33:49.196Z (12 months ago)
• Topics: attention, autoencoder, cloud-removal, generative-model, image-processing, image-restoration, lightweight, low-level-vision, remote-sensing
• Language: JavaScript
• Homepage: https://xavierjiezou.github.io/PMAA/
• Size: 32.5 MB
• Stars: 13
• Watchers: 3
• Forks: 2
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        


    

PMAA: A Progressive Multi-scale Attention Autoencoder Model for High-Performance Cloud Removal from Multi-temporal Satellite Imagery

This repository is the official PyTorch implementation of the accepted paper PMAA of ECAI 2023.





    Xuechao Zou1,*,

    Kai Li2,*,

    Junliang Xing2,

    Pin Tao1,2,†,

    Yachao Cui1





    Qinghai University1

    •

    Tsinghua University2

  

  


    Paper Preprint 

    |

    Project Page





![pmaa](image/README/pmaa.png)



## News

- [2023/07/30] Code release.

- [2023/07/16] PMAA got accepted by ECAI 2023.

- [2023/03/29] PMAA is on arXiv now.

## Requirements

To install dependencies:

```setup

pip install -r requirements.txt

```

To download datasets:

- _Sen2_MTC_Old_: [multipleImage.tar.gz](https://doi.org/10.7910/DVN/BSETKZ)

- _Sen2_MTC_New_: [CTGAN.zip](https://drive.google.com/file/d/1-hDX9ezWZI2OtiaGbE8RrKJkN1X-ZO1P/view?usp=share_link)

## Training

To train the models in the paper, run these commands:

```train

python train_old.py

python train_new.py

```

## Evaluation

To evaluate my models on two datasets, run:

```eval

python test_old.py

python test_new.py

```

## Pre-trained Models

You can download pretrained models here:

- Our awesome model trained on _Sen2_MTC_old_: [pmaa_old.pth](/pretrained/pmaa_old.pth)

- Our awesome model trained on _Sen2_MTC_new_: [pmaa_new.pth](/pretrained/pmaa_new.pth)

## Results

![res](image/README/res.png)

### Quantitative Results

![exp](image/README/exp.png)

### Qualitative Results

![vis](image/README/vis.png)

## Citation

If you use our code or models in your research, please cite with:

```latex

@article{zou2023pmaa,

  title={PMAA: A Progressive Multi-scale Attention Autoencoder Model for High-Performance Cloud Removal from Multi-temporal Satellite Imagery},

  author={Zou, Xuechao and Li, Kai and Xing, Junliang and Tao, Pin and Cui, Yachao},

  journal={European Conference on Artificial Intelligence (ECAI)},

  year={2023}

}

```