/adversarial-attacks-pytorch

PyTorch implementation of adversarial attacks.

Primary LanguagePythonMIT LicenseMIT

Adversarial-Attacks-PyTorch

MIT License Pypi Latest Release arXiv Documentation Status

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]).

import torchattacks
atk = torchattacks.PGD(model, eps=8/255, alpha=2/255, steps=4)
# If, images are normalized:
# atk.set_normalization_used(mean=[...], std=[...])
adv_images = atk(images, labels)

Table of Contents

  1. Requirements and Installation
  2. Getting Started
  3. Performance Comparison
  4. Citation
  5. Contribution
  6. Recommended Sites and Packages

Requirements and Installation

πŸ“‹ Requirements

  • PyTorch version >=1.4.0
  • Python version >=3.6

πŸ”¨ Installation from pip

pip install torchattacks

πŸ”¨ Installation from source (recommended)

pip install git+https://github.com/Harry24k/adversarial-attacks-pytorch.git

or

git clone https://github.com/Harry24k/adversarial-attacks-pytorch.git
cd adversarial-attacks-pytorch/
pip install -e .

Getting Started

⚠️ Precautions

  • All models should return ONLY ONE vector of (N, C) where C = number of classes. Considering most models in torchvision.models return one vector of (N,C), where N is the number of inputs and C is thenumber of classes, torchattacks also only supports limited forms of output. Please check the shape of the model’s output carefully.
  • 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 run torch.backends.cudnn.deterministic = True[ref].

πŸš€ Demos

Torchattacks supports following functions:

Targeted mode

  • Random target label:
# random labels as target labels.
atk.set_mode_targeted_random()
  • 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)
  • By label:
atk.set_mode_targeted_by_label(quiet=True)
# shift all class loops one to the right, 1=>2, 2=>3, .., 9=>0
target_labels = (labels + 1) % 10
adv_images = atk(images, target_labels)
  • Return to default:
atk.set_mode_default()

Save adversarial images

# Save
atk.save(data_loader, save_path="./data.pt", verbose=True)
  
# Load
adv_loader = atk.load(load_path="./data.pt")

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 search 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])

Torchattacks also supports collaboration with other attack packages.

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)

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)

πŸ”₯ List of implemented papers

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) 😍 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) 😍 Contributor taobai
TIFGSM
(Linf)
Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks (Dong et al., 2019) 😍 Contributor taobai
NIFGSM
(Linf)
Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks (Lin, et al., 2022) 😍 Contributor Zhijin-Ge
SINIFGSM
(Linf)
Nesterov Accelerated Gradient and Scale Invariance for Adversarial Attacks (Lin, et al., 2022) 😍 Contributor Zhijin-Ge
VMIFGSM
(Linf)
Enhancing the Transferability of Adversarial Attacks through Variance Tuning (Wang, et al., 2022) 😍 Contributor Zhijin-Ge
VNIFGSM
(Linf)
Enhancing the Transferability of Adversarial Attacks through Variance Tuning (Wang, et al., 2022) 😍 Contributor Zhijin-Ge
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)
LGV
(Linf, L2, L1, L0)
LGV: Boosting Adversarial Example Transferability from Large Geometric Vicinity (Gubri, et al., 2022) 😍 Contributor Martin Gubri
SPSA
(Linf)
Adversarial Risk and the Dangers of Evaluating Against Weak Attacks (Uesato, Jonathan, et al., 2018) 😍 Contributor Riko Naka
JSMA
(L0)
The Limitations of Deep Learning in Adversarial Settings (Papernot, Nicolas, et al., 2016) 😍 Contributor Riko Naka
EADL1
(L1)
EAD: Elastic-Net Attacks to Deep Neural Networks (Chen, Pin-Yu, et al., 2018) 😍 Contributor Riko Naka
EADEN
(L1, L2)
EAD: Elastic-Net Attacks to Deep Neural Networks (Chen, Pin-Yu, et al., 2018) 😍 Contributor Riko Naka
PIFGSM (PIM)
(Linf)
Patch-wise Attack for Fooling Deep Neural Network (Gao, Lianli, et al., 2020) 😍 Contributor Riko Naka
PIFGSM++ (PIM++)
(Linf)
Patch-wise++ Perturbation for Adversarial Targeted Attacks (Gao, Lianli, et al., 2021) 😍 Contributor Riko Naka

Performance Comparison

For a fair comparison, Robustbench is used. As for the comparison packages, currently updated and the most cited methods were selected:

  • Foolbox: 505 citations and last update 2022.10.
  • ART: 262 citations and last update 2022.10.

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) πŸ‘‘ ​Fastest
Foolbox* 0% (354ms) 44% (56ms) 58% (1856ms)
ART 0% (1384 ms) 44% (437ms) 58% (4704ms)
CW† (L2) Torchattacks 0% / 0.40
(2596ms)
14% / 0.61
(3795ms)
22% / 0.56
(43484ms)
πŸ‘‘ ​Highest Success Rate
πŸ‘‘ 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)
πŸ‘‘ 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.

†Considering that the binary search algorithm for const c can be time-consuming, torchattacks supports MutliAttack for grid searching c.

To push further, I introduce Rai-toolbox, which is newly added package!

Attack Package Time/step (accuracy)
FGSM (Linf) rai-toolbox 58 ms (0%)
Torchattacks 81 ms (0%)
Foolbox 105 ms (0%)
ART 83 ms (0%)
PGD (Linf) rai-toolbox 58 ms (44%)
Torchattacks 79 ms (44%)
Foolbox 82 ms (44%)
ART 90 ms (44%)
PGD (L2) rai-toolbox 58 ms (70%)
Torchattacks 81 ms (70%)
Foolbox 82 ms (70%)
ART 89 ms (70%)

The rai-toolbox takes a unique approach to gradient-based perturbations: they are implemented in terms of parameter-transforming optimizers and perturbation models. This enables users to implement diverse algorithms (like universal perturbations and concept probing with sparse gradients) using the same paradigm as a standard PGD attack.

Citation

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}
}

Recommended Sites and Packages