Awesome

FreMamba (IEEE TMM 2024)

📖Paper | 🖼️PDF

PyTorch codes for "Frequency-Assisted Mamba for Remote Sensing Image Super-Resolution", IEEE Transactions on Multimedia (TMM), 2024.

• Authors: Yi Xiao, Qiangqiang Yuan*, Kui Jiang, Yuzeng Chen, Qiang Zhang, and Chia-Wen Lin<br>

    <img src="fig/whu.png" width="110px">   <img src="fig/hit.png" width="150px">   <img src="fig/dmu.jpg" width="130px">   <img src="fig/nthu.png" width="150px">

🌱 Abstract

Recent progress in remote sensing image (RSI) super-resolution (SR) has exhibited remarkable performance using deep neural networks, e.g., Convolutional Neural Networks and Transformers. However, existing SR methods often suffer from either a limited receptive field or quadratic computational overhead, resulting in sub-optimal global representation and unacceptable computational costs in large-scale RSI. To alleviate these issues, we develop the first attempt to integrate the Vision State Space Model (Mamba) for RSI-SR, which specializes in processing large-scale RSI by capturing long-range dependency with linear complexity. To achieve better SR reconstruction, building upon Mamba, we devise a Frequency-assisted Mamba framework, dubbed FMSR, to explore the spatial and frequent correlations. In particular, our FMSR features a multi-level fusion architecture equipped with the Frequency Selection Module (FSM), Vision State Space Module (VSSM), and Hybrid Gate Module (HGM) to grasp their merits for effective spatial-frequency fusion. Recognizing that global and local dependencies are complementary and both beneficial for SR, we further recalibrate these multi-level features for accurate feature fusion via learnable scaling adaptors. Extensive experiments on AID, DOTA, and DIOR benchmarks demonstrate that our FMSR outperforms state-of-the-art Transformer-based methods HAT-L in terms of PSNR by 0.11 dB on average, while consuming only 28.05% and 19.08% of its memory consumption and complexity, respectively.

Overall

Install

git clone https://github.com/XY-boy/FreMamba.git

🎁 Dataset

Please download the following remote sensing benchmarks:

📃 Requirements

• CUDA 11.1
• Python 3.9.13
• PyTorch 1.9.1
• Torchvision 0.10.1
• causal_conv1d==1.0.0
• mamba_ssm==1.0.1

🧩 Usage

Test

• Step I. Use the structure below to prepare your dataset, e.g., DOTA, and DIOR. /xxxx/xxx/ (your data path)

/GT/ 
   /000.png  
   /···.png  
   /099.png  
/LR/ 
   /000.png  
   /···.png  
   /099.png  

• Step II. Change the --data_dir to your data path.
• Step III. Change the --pretrained_sr to your pre-trained model path.
• Step IV. Run the eval_4x.py

python eval_4x.py

Train

python train_4x.py

Acknowledgement

Our work is built upon MambaIR. Thanks to the author for sharing this awesome work!

🥰 Citation

If you find our work helpful in your research, please consider citing it!

@ARTICLE{xiao2024fmsr,
  author={Xiao, Yi and Yuan, Qiangqiang and Jiang, Kui and Chen, Yuzeng and Zhang, Qiang and Lin, Chia-Wen},
  journal={IEEE Transactions on Multimedia}, 
  title={Frequency-Assisted Mamba for Remote Sensing Image Super-Resolution}, 
  year={2024},
  volume={26},
  number={},
  pages={1-14},
}