Skip to content

Latest commit

 

History

History
36 lines (28 loc) · 1.61 KB

README.md

File metadata and controls

36 lines (28 loc) · 1.61 KB

RCCA

This repository contains the PyTorch implementation of the paper "A New Context-Aware Framework for Defending Against Adversarial Attacks in Hyperspectral Image Classification". (IEEE TGRS 2023) [paper]
By Bing Tu, Wangquan He, Qianming Li, Yishu Peng, and Antonio Plaza

Usage

1. Installation

Requirements are Python 3.7 and PyTorch 1.4.0.

2. Download data

Download the dataset from [here], and put it in the 'Data' directory.

3. Training and test

To train the model with different attack strategies, please run the following command:

python demo_RCCA_FGSM.py --attack FGSM

In addition, the dataset and model can be replaced by changing --dataID and --model in 'demo_RCCA_FGSM.py'.

Citation

Please consider citing our work if you think it is useful for your research：

@ARTICLE{tu2023new,
  author={Tu, Bing and He, Wangquan and Li, Qianming and Peng, Yishu and Plaza, Antonio},
  journal={IEEE Trans. Geosci. Remote Sens.}, 
  title={A New Context-Aware Framework for Defending Against Adversarial Attacks in Hyperspectral Image Classification}, 
  year={2023},
  volume={61},
  month={Feb},
  pages={1-14},
  doi={10.1109/TGRS.2023.3250450}}

Acknowledgment: This code is based on the SACNet and UAE-RS. Thanks to the authors for their wonderful work.