Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methods — mentioned by tMcGrath

Thu, 02 Nov 2017 07:57:57 GMT

I agree - it's a surprising and cool paper. There has been some work done on fooling network ensembles by constructing constructing a Bayesian posterior over weights using dropout [0]. This is an ensemble of weights for the same network, not over different architectures, however. The basic idea here is that most of the time, each member of the ensemble will misclassify the adversarial example in a different way. This means that the posterior predictive distribution for adversarial examples ends up much broader, and you can detect them this way. Surprisingly, even this can be beaten in the…

Paper and discussion history. A mention is not an endorsement.

Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methods · HackerPapers discussions

Adversarial Examples Are Not Easily Detected: Bypassing Ten Detection Methods — mentioned by tMcGrath