sparse-imperceivable-attacks

sparse-imperceivable-attacks copied to clipboard

# cornersearch_attacks_pt.py # line 229 #print('attack successful: {:.2f}%'.format((1.0 - np.mean(fl_success))*100.0)) I think this statement should be changed like this -> #print('attack successful: {:.2f}%'.format((np.mean(fl_success))*100.0))

wildbrother

how to run and test on ImageNet?

1

Hi I have ImageNet samples and did try to test your method on ImageNet but it seems that it causes out of memory. How did you address this issue? ```...

vietvo89