>For the chief reason that there's probably nothing better. The paper "Deep Neural Networks are Easily Fooled" noted that generative models don't suffer from t

>For the chief reason that there's probably nothing better. The paper "Deep Neural Networks are Easily Fooled" noted that generative models don't suffer from this flaw.