-
In the # 105 of deeprobust/image/attack/deepfool.py. The code was change from zero_gradients(x) to x.zero_grad(), which will cause the "AttributeError: 'Tensor' object has no attribute 'zero_grad'" er…
-
As shown in title, I use `torchattacks.DeepFool(model)` to generate attack data about cifar100 trainset, batch size is 512 and had run in Tesla P40 20 hours. But no results, is it normally? here is c…
-
I think this assert is confused for user
https://github.com/Harry24k/adversarial-attacks-pytorch/blob/6dbe9155b0ba6ff966f2d484366c13fcbf80e38d/torchattacks/attacks/pixle.py#L83-L84
after that, u…
-
Why I cannot pass steps or overshoot parameters to deepfool attack??
same happens for CWL2 attacks, i cannot pass steps or override parameters?
whay I should pass epsilon to deepfool and to CWL2?
T…
-
I have some doubts and hope to get your reply.
1. The `images` obtained by `ep.astensors(samples(fmodel, dataset="imagenet", batchsize=5))` are only normalized by 255, i think the operation `(image…
-
Hi, I was confused by the results of L2 DeepFool Attacks and L2 CW Attacks. I tested both, their accuracies are 89%, whatever I changed. The changing includes steps, overshoot, stepsize, and epsilons.…
-
For me, the Deepfool attack is not working as expected.
While I'm getting good results for FGSM and BIM, Deepfool only outputs the Gradient of the image.
The bounds + preprocessing is exactly the …
-
Referring to this [Colab Notebook](https://colab.research.google.com/github/jonasrauber/foolbox-native-tutorial/blob/master/foolbox-native-tutorial.ipynb).
I have just replaced the L_Inf DeepFool a…
-
This was reported to me by @max-andr. Most of the differences are actually explicitly mentioned in comments in our implementation, but we should check again if we can match the reference implementatio…
-
This was reported to me by @max-andr.
Possibly, changing the `residual_labels` function to use `