English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Recent works have demonstrated the existence of {\it adversarial examples}targeting a single machine learning system. In this paper we ask a simple butfundamental question of "selective fooling": given {\it multiple} machinelearning systems assigned to solve the same classification problem and takingthe same input signal, is it possible to construct a perturbation to the inputsignal that manipulates the outputs of these {\it multiple} machine learningsystems {\it simultaneously} in arbitrary pre-defined ways? For example, is itpossible to selectively fool a set of "enemy" machine learning systems but doesnot fool the other "friend" machine learning systems? The answer to thisquestion depends on the extent to which these different machine learningsystems "think alike". We formulate the problem of "selective fooling" as anovel optimization problem, and report on a series of experiments on the MNISTdataset. Our preliminary findings from these experiments show that it is infact very easy to selectively manipulate multiple MNIST classifierssimultaneously, even when the classifiers are identical in their architectures,training algorithms and training datasets except for random initializationduring training. This suggests that two nominally equivalent machine learningsystems do not in fact "think alike" at all, and opens the possibility for manynovel applications and deeper understandings of the working principles of deepneural networks.

Do Deep Minds Think Alike? Selective Adversarial Attacks for Fine-Grained Manipulation of Multiple Deep Neural Networks