4 ms·
This is a fascinating paper. Does training on generated noisy images make the resultant classifier more resistant to standard adversarial examples?
by ravel-bar-foo 5y ago
This is a fascinating paper. Does training on generated noisy images make the resultant classifier more resistant to standard adversarial examples?