On the Effectiveness of Defensive Distillation
نویسندگان
چکیده
We report experimental results indicating that defensive distillation successfully mitigates adversarial samples crafted using the fast gradient sign method [2], in addition to those crafted using the Jacobian-based iterative attack [5] on which the defense mechanism was originally evaluated.
منابع مشابه
Defensive Distillation is Not Robust to Adversarial Examples
We show that defensive distillation is not secure: it is no more resistant to targeted misclassification attacks than unprotected neural networks.
متن کاملExtending Defensive Distillation
Machine learning is vulnerable to adversarial examples: inputs carefully modified to force misclassification. Designing defenses against such inputs remains largely an open problem. In this work, we revisit defensive distillation—which is one of the mechanisms proposed to mitigate adversarial examples—to address its limitations. We view our results not only as an effective way of addressing som...
متن کاملEnhanced Attacks on Defensively Distilled Deep Neural Networks
Deep neural networks (DNNs) have achieved tremendous success in many tasks of machine learning, such as the image classification. Unfortunately, researchers have shown that DNNs are easily attacked by adversarial examples, slightly perturbed images which can mislead DNNs to give incorrect classification results. Such attack has seriously hampered the deployment of DNN systems in areas where sec...
متن کاملInvestigating the Components of Defensive Culture in the Strategic Culture Framework of I.R.Iran (Based on Ayatollah Khamenei's views)
If culture is seen as a framework of "should" and "should not", defensive culture isbased on issues, which are availablein the society. Issues that differ from one society to another and are influenced by the strategic culture of the countries. So, the present study aimed to answer the question that what are the most important components of a defensive culture in the framework of the strategic ...
متن کاملOn global (strong) defensive alliances in some product graphs
A defensive alliance in a graph is a set $S$ of vertices with the property that every vertex in $S$ has at most one moreneighbor outside of $S$ than it has inside of $S$. A defensive alliance $S$ is called global if it forms a dominating set. The global defensive alliance number of a graph $G$ is the minimum cardinality of a global defensive alliance in $G$. In this article we study the global ...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- CoRR
دوره abs/1607.05113 شماره
صفحات -
تاریخ انتشار 2016