Torchattacks is a PyTorch library that provides adversarial attacks to generate adversarial examples. It contains PyTorch-like interface and functions that make it easier for PyTorch users to implement adversarial attacks (README [KOR]).
Easy implementation
import torchattacks
atk = torchattacks.PGD(model, eps=8/255, alpha=2/255, steps=4)
adv_images = atk(images, labels)
Easy modification
from torchattacks.attack import Attack
class CustomAttack(Attack):
def __init__(self, model):
super().__init__("CustomAttack", model)
def forward(self, images, labels=None):
adv_images = # Custom attack method
return adv_images
Useful functions
atk.set_mode_targeted_least_likely(kth_min) # Targeted attack
atk.set_return_type(type='int') # Return values [0, 255]
atk = torchattacks.MultiAttack([atk1, ..., atk99]) # Combine attacks
atk.save(data_loader, save_path=None, verbose=True, return_verbose=False) # Save adversarial images
- Requirements and Installation
- Getting Started
- Performance Comparison
- Citation
- Contribution
- Recommended Sites and Packages
- PyTorch version >=1.4.0
- Python version >=3.6
pip install torchattacks
- All images should be scaled to [0, 1] with transform[to.Tensor()] before used in attacks. To make it easy to use adversarial attacks, a reverse-normalization is not included in the attack process. To apply an input normalization, please add a normalization layer to the model. Please refer to code or nbviewer.
- All models should return ONLY ONE vector of
(N, C)
whereC = number of classes
. Considering most models in torchvision.models return one vector of(N,C)
, whereN
is the number of inputs andC
is thenumber of classes, torchattacks also only supports limited forms of output. Please check the shape of the modelβs output carefully. In the case of the model returns multiple outputs, please refer to the demo. torch.backends.cudnn.deterministic = True
to get same adversarial examples with fixed random seed. Some operations are non-deterministic with float tensors on GPU [discuss]. If you want to get same results with same inputs, please runtorch.backends.cudnn.deterministic = True
[ref].
import torchattacks
atk = torchattacks.PGD(model, eps=8/255, alpha=2/255, steps=4)
adv_images = atk(images, labels)
Targeted mode
- Random target label:
# random labels as target labels.
atk.set_mode_targeted_random(n_classses)
- Least likely label:
# label with the k-th smallest probability used as target labels.
atk.set_mode_targeted_least_likely(kth_min)
- By custom function:
# label from mapping function
atk.set_mode_targeted_by_function(target_map_function=lambda images, labels:(labels+1)%10)
- Return to default:
atk.set_mode_default()
Return type
- Return adversarial images with integer value (0-255).
atk.set_return_type(type='int')
- Return adversarial images with float value (0-1).
atk.set_return_type(type='float')
Save adversarial images
# Save
atk.save(data_loader, save_path="./data/sample.pt", verbose=True)
# Load
import torch
from torch.utils.data import DataLoader, TensorDataset
adv_images, labels = torch.load("./data/sample.pt")
# If set_return_type was 'int',
# adv_data = TensorDataset(adv_images.float()/255, labels)
# else,
adv_data = TensorDataset(adv_images, labels)
adv_loader = DataLoader(adv_data, batch_size=128, shuffle=False)
Training/Eval during attack
# For RNN-based models, we cannot calculate gradients with eval mode.
# Thus, it should be changed to the training mode during the attack.
atk.set_training_mode(model_training=False, batchnorm_training=False, dropout_training=False)
Make a set of attacks
- Strong attacks
atk1 = torchattacks.FGSM(model, eps=8/255)
atk2 = torchattacks.PGD(model, eps=8/255, alpha=2/255, iters=40, random_start=True)
atk = torchattacks.MultiAttack([atk1, atk2])
- Binary serach for CW
atk1 = torchattacks.CW(model, c=0.1, steps=1000, lr=0.01)
atk2 = torchattacks.CW(model, c=1, steps=1000, lr=0.01)
atk = torchattacks.MultiAttack([atk1, atk2])
- Random restarts
atk1 = torchattacks.PGD(model, eps=8/255, alpha=2/255, iters=40, random_start=True)
atk2 = torchattacks.PGD(model, eps=8/255, alpha=2/255, iters=40, random_start=True)
atk = torchattacks.MultiAttack([atk1, atk2])
- White Box Attack with ImageNet (code, nbviewer): Using torchattacks to make adversarial examples with the ImageNet dataset to fool ResNet-18.
- Transfer Attack with CIFAR10 (code, nbviewer): This demo provides an example of black box attack with two different models. First, make adversarial datasets from a holdout model with CIFAR10 and save it as torch dataset. Second, use the adversarial datasets to attack a target model.
- Adversairal Training with MNIST (code, nbviewer): This code shows how to do adversarial training with this repository. The MNIST dataset and a custom model are used in this code. The adversarial training is performed with PGD, and then FGSM is applied to evaluate the model.
FoolBox
https://github.com/bethgelab/foolbox
from torchattacks.attack import Attack
import foolbox as fb
# L2BrendelBethge
class L2BrendelBethge(Attack):
def __init__(self, model):
super(L2BrendelBethge, self).__init__("L2BrendelBethge", model)
self.fmodel = fb.PyTorchModel(self.model, bounds=(0,1), device=self.device)
self.init_attack = fb.attacks.DatasetAttack()
self.adversary = fb.attacks.L2BrendelBethgeAttack(init_attack=self.init_attack)
self._attack_mode = 'only_default'
def forward(self, images, labels):
images, labels = images.to(self.device), labels.to(self.device)
# DatasetAttack
batch_size = len(images)
batches = [(images[:batch_size//2], labels[:batch_size//2]),
(images[batch_size//2:], labels[batch_size//2:])]
self.init_attack.feed(model=self.fmodel, inputs=batches[0][0]) # feed 1st batch of inputs
self.init_attack.feed(model=self.fmodel, inputs=batches[1][0]) # feed 2nd batch of inputs
criterion = fb.Misclassification(labels)
init_advs = self.init_attack.run(self.fmodel, images, criterion)
# L2BrendelBethge
adv_images = self.adversary.run(self.fmodel, images, labels, starting_points=init_advs)
return adv_images
atk = L2BrendelBethge(model)
atk.save(data_loader=test_loader, save_path="_temp.pt", verbose=True)
Adversarial-Robustness-Toolbox (ART)
https://github.com/IBM/adversarial-robustness-toolbox
import torch.nn as nn
import torch.optim as optim
from torchattacks.attack import Attack
import art.attacks.evasion as evasion
from art.classifiers import PyTorchClassifier
# SaliencyMapMethod (or Jacobian based saliency map attack)
class JSMA(Attack):
def __init__(self, model, theta=1/255, gamma=0.15, batch_size=128):
super(JSMA, self).__init__("JSMA", model)
self.classifier = PyTorchClassifier(
model=self.model, clip_values=(0, 1),
loss=nn.CrossEntropyLoss(),
optimizer=optim.Adam(self.model.parameters(), lr=0.01),
input_shape=(1, 28, 28), nb_classes=10)
self.adversary = evasion.SaliencyMapMethod(classifier=self.classifier,
theta=theta, gamma=gamma,
batch_size=batch_size)
self.target_map_function = lambda labels: (labels+1)%10
self._attack_mode = 'only_default'
def forward(self, images, labels):
adv_images = self.adversary.generate(images, self.target_map_function(labels))
return torch.tensor(adv_images).to(self.device)
atk = JSMA(model)
atk.save(data_loader=test_loader, save_path="_temp.pt", verbose=True)
The distance measure in parentheses.
Name | Paper | Remark |
---|---|---|
FGSM (Linf) |
Explaining and harnessing adversarial examples (Goodfellow et al., 2014) | |
BIM (Linf) |
Adversarial Examples in the Physical World (Kurakin et al., 2016) | Basic iterative method or Iterative-FSGM |
CW (L2) |
Towards Evaluating the Robustness of Neural Networks (Carlini et al., 2016) | |
RFGSM (Linf) |
Ensemble Adversarial Traning: Attacks and Defences (Tramèr et al., 2017) | Random initialization + FGSM |
PGD (Linf) |
Towards Deep Learning Models Resistant to Adversarial Attacks (Mardry et al., 2017) | Projected Gradient Method |
PGDL2 (L2) |
Towards Deep Learning Models Resistant to Adversarial Attacks (Mardry et al., 2017) | Projected Gradient Method |
MIFGSM (Linf) |
Boosting Adversarial Attacks with Momentum (Dong et al., 2017) | π Contributor zhuangzi926, huitailangyz |
TPGD (Linf) |
Theoretically Principled Trade-off between Robustness and Accuracy (Zhang et al., 2019) | |
EOTPGD (Linf) |
Comment on "Adv-BNN: Improved Adversarial Defense through Robust Bayesian Neural Network" (Zimmermann, 2019) | EOT+PGD |
APGD (Linf, L2) |
Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks (Croce et al., 2020) | |
APGDT (Linf, L2) |
Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks (Croce et al., 2020) | Targeted APGD |
FAB (Linf, L2, L1) |
Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack (Croce et al., 2019) | |
Square (Linf, L2) |
Square Attack: a query-efficient black-box adversarial attack via random search (Andriushchenko et al., 2019) | |
AutoAttack (Linf, L2) |
Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks (Croce et al., 2020) | APGD+APGDT+FAB+Square |
DeepFool (L2) |
DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks (Moosavi-Dezfooli et al., 2016) | |
OnePixel (L0) |
One pixel attack for fooling deep neural networks (Su et al., 2019) | |
SparseFool (L0) |
SparseFool: a few pixels make a big difference (Modas et al., 2019) | |
DIFGSM (Linf) |
Improving Transferability of Adversarial Examples with Input Diversity (Xie et al., 2019) | π Contributor taobai |
TIFGSM (Linf) |
Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks (Dong et al., 2019) | π Contributor taobai |
Jitter (Linf) |
Exploring Misclassifications of Robust Neural Networks to Enhance Adversarial Attacks (Schwinn, Leo, et al., 2021) | |
Pixle (L0) |
Pixle: a fast and effective black-box attack based on rearranging pixels (Pomponi, Jary, et al., 2022) |
For a fair comparison, Robustbench is used. As for the comparison packages, currently updated and the most cited methods were selected:
Robust accuracy against each attack and elapsed time on the first 50 images of CIFAR10. For L2 attacks, the average L2 distances between adversarial images and the original images are recorded. All experiments were done on GeForce RTX 2080. For the latest version, please refer to here (code, nbviewer).
Attack | Package | Standard | Wong2020Fast | Rice2020Overfitting | Remark |
---|---|---|---|---|---|
FGSM (Linf) | Torchattacks | 34% (54ms) | 48% (5ms) | 62% (82ms) | |
Foolbox* | 34% (15ms) | 48% (8ms) | 62% (30ms) | ||
ART | 34% (214ms) | 48% (59ms) | 62% (768ms) | ||
PGD (Linf) | Torchattacks | 0% (174ms) | 44% (52ms) | 58% (1348ms) | π βFastest |
Foolbox* | 0% (354ms) | 44% (56ms) | 58% (1856ms) | ||
ART | 0% (1384 ms) | 44% (437ms) | 58% (4704ms) | ||
CWβ Β (L2) | Torchattacks | 0% / 0.40 (2596ms) |
14% / 0.61 (3795ms) |
22% / 0.56 (43484ms) |
π βHighest Success Rate π Fastest |
Foolbox* | 0% / 0.40 (2668ms) |
32% / 0.41 (3928ms) |
34% / 0.43 (44418ms) |
||
ART | 0% / 0.59 (196738ms) |
24% / 0.70 (66067ms) |
26% / 0.65 (694972ms) |
||
PGD (L2) | Torchattacks | 0% / 0.41 (184ms) | 68% / 0.5 (52ms) |
70% / 0.5 (1377ms) |
π Fastest |
Foolbox* | 0% / 0.41 (396ms) | 68% / 0.5 (57ms) |
70% / 0.5 (1968ms) |
||
ART | 0% / 0.40 (1364ms) | 68% / 0.5 (429ms) |
70% / 0.5 (4777ms) |
* Note that Foolbox returns accuracy and adversarial images simultaneously, thus the actual time for generating adversarial images might be shorter than the records.
β Considering that the binary search algorithm for const c
can be time-consuming, torchattacks supports MutliAttack for grid searching c
.
If you use this package, please cite the following BibTex (SemanticScholar, GoogleScholar):
@article{kim2020torchattacks,
title={Torchattacks: A pytorch repository for adversarial attacks},
author={Kim, Hoki},
journal={arXiv preprint arXiv:2010.01950},
year={2020}
}
All kind of contributions are always welcome! π
If you are interested in adding a new attack to this repo or fixing some issues, please have a look at CONTRIBUTING.md.
-
Adversarial Attack Packages:
- https://github.com/IBM/adversarial-robustness-toolbox: Adversarial attack and defense package made by IBM. TensorFlow, Keras, Pyotrch available.
- https://github.com/bethgelab/foolbox: Adversarial attack package made by Bethge Lab. TensorFlow, Pyotrch available.
- https://github.com/tensorflow/cleverhans: Adversarial attack package made by Google Brain. TensorFlow available.
- https://github.com/BorealisAI/advertorch: Adversarial attack package made by BorealisAI. Pytorch available.
- https://github.com/DSE-MSU/DeepRobust: Adversarial attack (especially on GNN) package made by BorealisAI. Pytorch available.
- https://github.com/fra31/auto-attack: Set of attacks that is believed to be the strongest in existence. TensorFlow, Pyotrch available.
-
Adversarial Defense Leaderboard:
-
Adversarial Attack and Defense Papers:
- https://nicholas.carlini.com/writing/2019/all-adversarial-example-papers.html: A Complete List of All (arXiv) Adversarial Example Papers made by Nicholas Carlini.
- https://github.com/chawins/Adversarial-Examples-Reading-List: Adversarial Examples Reading List made by Chawin Sitawarin.
-
ETC:
- https://github.com/Harry24k/gnn-meta-attack: Adversarial Poisoning Attack on Graph Neural Network.
- https://github.com/ChandlerBang/awesome-graph-attack-papers: Graph Neural Network Attack papers.