Foolbox is a Python toolbox to create adversarial examples that fool neural networks. It requires Python, NumPy and SciPy.
Foolbox is maintained and supported here
This is just where I post attacks I need for testing.
Same installation procedure.
pip install foolboxThough it is highly recommended that you install in editable mode
git clone https://github.com/neale/foolbox.git
cd foolbox
pip install --editable .I'm testing with Python 2.7 here. I don't want to be, but here we are. That said, everything should by Python 3.5 complient.
The real documentation is available here
I have posted a thorough example with each new attack, as well at the attack itself at the top level directory.