Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Problem about CW-100 #1

Open
menghaoma opened this issue Nov 28, 2019 · 1 comment
Open

Problem about CW-100 #1

menghaoma opened this issue Nov 28, 2019 · 1 comment

Comments

@menghaoma
Copy link

Hi Alan,
Thanks for your code. When I applied CW-100 to attack, I got a defense result of 0%. The code is：

adversary = CarliniWagnerL2Attack(net,num_classes=10,targeted=False,max_iterations=args.iteration)

I am wondering if there is any problem using CW-100 attack like this？
Thanks again for your work.
@AlanChou
Copy link
Owner

Hi,

I just tested it with the adversary you mentioned and I also got 0%. I'm currently busy and don't have the time to find the problem and fix it in the near future. I'll suggest you to use the same test script to test other's implementation as I noticed that there are several implementations of "Adversarial training for free".

Please let me know if you find out that there's anything wrong. I'll leave this issue open till it's fixed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@AlanChou @menghaoma