Abstract: Recent researches demonstrate that deep learning models are vulnerable to membership inference attacks. Few defenses have been proposed, but suffer from compromising the performance or ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback