Skip to content

CVPR 2019 Paper——Feature Space Perturbations Yield More Transferable Adversarial Examples re-implementation.

License

Notifications You must be signed in to change notification settings

QwQ2000/Activation-Attack-Pytorch

Repository files navigation

Activation-Attack-Pytorch

Apply targeted attacks on black box ViT with the activation attack method. The white box model is ResNet-18.

Based on CVPR 2019 Paper Feature Space Perturbations Yield More Transferable Adversarial Examples.

My additional expreriment proves that the transferablitiy is low between heterogeneous deep models.

All the models are finetuned on CIFAR-10 with the Adam optimizer with lr = 1e-5.

Pretrained models

Model Name CIFAR-10 Test Accuracy ImageNet-1K Pretrain
ResNet-18 0.9435 Yes
DenseNet-121 0.9571 Yes
ViT-B/16 0.9838 Yes

Results

Model Error uTr tSuc tTr
ResNet-18 96.39 / 75.48 /
DenseNet-121 80.16 82.27 9.39 10.65
ViT 22.33 22.78 2.5 4.2

About

CVPR 2019 Paper——Feature Space Perturbations Yield More Transferable Adversarial Examples re-implementation.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages