Python toolbox to create adversarial examples that fool neural networks
Project description
Foolbox
Foolbox is a Python toolbox to create adversarial examples that fool neural networks. It requires Python, NumPy and SciPy.
Installation
# Foolbox 1.8
pip install foolbox
# Foolbox 2.0 release candidate
pip install foolbox --pre
Foolbox requires Python 3.5 or newer (since Foolbox 2.0).
Documentation
Documentation for the latest stable version as well as pre-release versions is available on ReadTheDocs.
Our paper describing Foolbox is on arXiv: https://arxiv.org/abs/1707.04131
Example
import foolbox
import keras
import numpy as np
from keras.applications.resnet50 import ResNet50
# instantiate model
keras.backend.set_learning_phase(0)
kmodel = ResNet50(weights='imagenet')
preprocessing = (np.array([104, 116, 123]), 1)
fmodel = foolbox.models.KerasModel(kmodel, bounds=(0, 255), preprocessing=preprocessing)
# get source image and label
image, label = foolbox.utils.imagenet_example()
# apply attack on source image
# ::-1 reverses the color channels, because Keras ResNet50 expects BGR instead of RGB
attack = foolbox.attacks.FGSM(fmodel)
adversarial = attack(image[:, :, ::-1], label)
# if the attack fails, adversarial will be None and a warning will be printed
For more examples, have a look at the documentation.
Finally, the result can be plotted like this:
# if you use Jupyter notebooks
%matplotlib inline
import matplotlib.pyplot as plt
plt.figure()
plt.subplot(1, 3, 1)
plt.title('Original')
plt.imshow(image / 255) # division by 255 to convert [0, 255] to [0, 1]
plt.axis('off')
plt.subplot(1, 3, 2)
plt.title('Adversarial')
plt.imshow(adversarial[:, :, ::-1] / 255) # ::-1 to convert BGR to RGB
plt.axis('off')
plt.subplot(1, 3, 3)
plt.title('Difference')
difference = adversarial[:, :, ::-1] - image
plt.imshow(difference / abs(difference).max() * 0.2 + 0.5)
plt.axis('off')
plt.show()
Interfaces for a range of other deeplearning packages such as TensorFlow 1 and 2, PyTorch, Theano, Lasagne and MXNet are available, e.g.
model = foolbox.models.TensorFlowModel(images, logits, bounds=(0, 255))
model = foolbox.models.TensorFlowEagerModel(model, bounds=(0, 255))
model = foolbox.models.PyTorchModel(torchmodel, bounds=(0, 255), num_classes=1000)
# etc.
Different adversarial criteria such as Top-k, specific target classes or target probability values for the original class or the target class can be passed to the attack, e.g.
criterion = foolbox.criteria.TargetClass(22)
attack = foolbox.attacks.LBFGSAttack(fmodel, criterion)
Feature requests and bug reports
We welcome feature requests and bug reports. Just create a new issue on GitHub.
Questions & FAQ
Depending on the nature of your question feel free to post it as an issue on GitHub, or post it as a question on Stack Overflow using the foolbox tag. We will try to monitor that tag but if you don’t get an answer don’t hesitate to contact us.
Before you post a question, please check our FAQ and our Documentation on ReadTheDocs.
Contributions welcome
Foolbox is a work in progress and any input is welcome. Foolbox is particularly well-suited to develop new adversarial attacks and to support new machine learning and deep learning frameworks by simply adding a wrapper. By adding reference implementations for adversarial attacks to Foolbox, they will automatically be applicable to models implemented in any of the supported frameworks such as PyTorch, TensorFlow, Keras Or MxNet.
Citation
If you use Foolbox for your work, please cite our paper:
@article{rauber2017foolbox,
title={Foolbox: A Python toolbox to benchmark the robustness of machine learning models},
author={Rauber, Jonas and Brendel, Wieland and Bethge, Matthias},
journal={arXiv preprint arXiv:1707.04131},
year={2017},
url={http://arxiv.org/abs/1707.04131},
archivePrefix={arXiv},
eprint={1707.04131},
}
You can find the paper on arXiv: https://arxiv.org/abs/1707.04131
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.