1 paper · 1 filter
Peter Blohm, Patrick Indri, Thomas Gärtner +1
We propose and investigate probabilistic guarantees for the adversarial robustness of classification algorithms. While traditional formal verification approaches for robustness are…