IEEE TGRS---MS-CVNets-a-novel-SAR-ATR-method-based-on-Multi-Stream-Complex-Valued-Networks

The code in this toolbox implements the "SAR Automatic Target Recognition Method based on Muliti-Stream Complex-Valued Networks" in IEEE Transactions on Geoscience and Remote Sensing (TGRS). More specifically, it is detailed as follow.

usage

Complex-MSTAR dataset

this Complex-MSTAR dataset is based on the original MSTAR program. we do not participate in the data acquisition work, only data redistribution and collation. the complex-mstar dataset structure is as follows:

---Complex-MSTAR-
-----------------data_SOC-class10-trian-imgs
----------------------------------------data_train_64.mat
----------------------------------------data_train_128.mat
----------------------------------test-data_test_64.mat
---------------------------------------data_test_128.mat
-------------------------calss3-trian--data_train_64.mat
---------------------------------------data_train_128.mat
----------------------------------test-data_test_64.mat
---------------------------------------data_test_128.mat
-----------------data_EOC-depression_variation
------------------------ -noise varision
--------------------------version variation
the Complex-MSTAR dataset provide both size of 64x64 and 128x128 for different task requriements.

requriements

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

Training

•         python Train.py                        
  

Testing

•          python Test.py                 
  

Citation

please kindly cite this paper if our MS-CVNets can give you any inspiration for your research, thanks a lot.

Z. Zeng, J. Sun, Z. Han and W. Hong, "SAR Automatic Target Recognition Method Based on Multi-Stream Complex-Valued Networks," in IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-18, 2022, Art no. 5228618, doi: 10.1109/TGRS.2022.3177323.

Contact

Zhiqiang Zeng
Email:zengzq@buaa.edu.cn

references

1. the original MSTAR dataset information: https://www.sdms.afrl.af.mil/
2. we would like to appreciate the ChihebTrabelsi, wavefrontshaping and ivannz in the help of coding basis. we bulid the MS-CVNets principal framework on their basis.
   
3. https://github.com/ivannz/cplxmodule
4. https://github.com/wavefrontshaping/complexPyTorch
5. https://github.com/ChihebTrabelsi/deep_complex_networks