no code implementations • 17 Feb 2023 • David Weber, Florian Merkle, Pascal Schöttle, Stephan Schlögl
To do so, we conducted a pre-study and two human-grounded experiments, assessing the effects of different pruning ratios on CNN explainability.
no code implementations • 16 Feb 2023 • Roland Rauter, Martin Nocker, Florian Merkle, Pascal Schöttle
Another type of adversarial examples are invariance-based adversarial examples, where the images are semantically modified such that the predicted class of the model does not change, but the class that is determined by humans does.
1 code implementation • 16 Feb 2023 • Martin Nocker, David Drexel, Michael Rader, Alessio Montuoro, Pascal Schöttle
Fully homomorphic encryption (FHE) is a promising technique to enable individuals using ML services without giving up privacy and protecting the ML model of service providers at the same time.
no code implementations • 19 Aug 2021 • Florian Merkle, Maximilian Samsinger, Pascal Schöttle
Available research on the impact of neural network pruning on the adversarial robustness is fragmentary and often does not adhere to established principles of robustness evaluation.
no code implementations • 17 Aug 2021 • Maximilian Samsinger, Florian Merkle, Pascal Schöttle, Tomas Pevny
Adversarial machine learning, i. e., increasing the robustness of machine learning algorithms against so-called adversarial examples, is now an established field.
no code implementations • 7 Feb 2020 • Thomas Baumhauer, Pascal Schöttle, Matthias Zeppelzauer
Recently enacted legislation grants individuals certain rights to decide in what fashion their personal data may be used, and in particular a "right to be forgotten".