a novel SAR-ATR model by considering both CNN-based features and transformer-based features to further enhance the representation ability of SAR targets under limited samples.

This is a PyTorch implementation of the "Multi-Level Attention Networks for Synthetic Aperture Radar Automatic Target Recognition" in IEEE Geoscience and Remote Sensing Letters (GRSL). More specifically, it is detailed as follow.

---pytorch 1.6
---python 3.7
---CUDA 10.1

•  python train.py
  

•  python test.py
  

e-mail: zengzq@buaa.edu.cn

Please kindly cite the papers if this code is useful and helpful for your research.
Y. Guo, Z. Zeng, M. Jin, J. Sun, Z. Meng and W. Hong, "Multi-Level Attention Networks for Synthetic Aperture Radar Automatic Target Recognition," in IEEE Geoscience and Remote Sensing Letters, doi: 10.1109/LGRS.2024.3417222.