-
With the current implementation, every object created from the classes based on the `Attack` class need to provide a `tf.Session` instance in the constructor. However, this limits the usage of the att…
-
I was trying deepfool attack on Mnist. For the 10,000 test images, the attack returned around 9500 of them, but not all of them are adversarial as about for 30% of these the models predictions is same…
a7b23 updated
6 years ago
-
-
My model successfully generates adversarial images on LGBFS but get None with the same code on deep fool attack.
Through debugging, I found really strange behavior
```
adv = attack(x, y) // Got N…
-
Foolbox tensorflow model predictions donot match tensorflow slim model predictions on clean images .
Is there some issue while restoring the model ?
```
from __future__ import absolute_import
…
-
How create training samples from adversarially perturbed original training sampels?
To be simple, suppose I had 100 training images and wanted to use Deep Fool and FGSM to perturb these samples, I …
pGit1 updated
6 years ago
-
It seems that the perturbations computed by the FastGradientMethod (FGM) and BasicIterativeMethod attacks are sometimes equal to zero.
In a first preliminary test (not in cleverhans), this problem …
ricvo updated
6 years ago
-
I found an issue when running the python test_deepfool code. Could be my machine, but I this is what I ran into with a fresh pytorch 0.1.12
The test file doesn't work right away for me, I had to chan…
neale updated
7 years ago
-
The reference implementation (Python, PyTorch) is available at
https://github.com/LTS4/DeepFool/tree/master/Python
-
in this line https://github.com/bethgelab/foolbox/blob/master/foolbox/adversarial.py#L226, you check whether the adversarial image is within bound or not, but we should add the offset back first. C…
Ailss updated
6 years ago