SparseAG_GS

An efficient and sparse adversarial test case generation method for CV(computer Vision) software

Code release and supplementary materials for：

"SparseAG-GS: Adversarial Test Case Generation via Sparse Perturbation Group"

Datasets

• CIFAR-10
• ImageNet

Dependencies

The code was tested with:

• h5py 3.8.0
• ipykernel 6.19.2
• matplotlib 3.7.2
• numpy 1.25.2
• pandas 1.5.3
• scikit-image 0.21.0
• scipy 1.9.3
• torch 1.11.0
• torchvision 0.12.0
• tqdm 4.64.1

Training

Training the improved AdvGAN for generating the importance matrix of perturbations.

python train_advGAN.py

Evaluations

1. Non-target attacks on CIFAR-10

   python SparseAG_GS_cifar.py
   

2. Target attacks on CIFAR-10

   python SparseAG_GS_Tarcifar.py
   

3. Non-target attacks on ImageNet

   python SparseAG_GS_imagenet.py
   

4. Target attacks on ImageNet

   python SparseAG_GS_imagenetTar.py 
   

5. Ablation study

   python aeCifar.py