Vol-4141⫷ Vol-4142 ⫸Vol-4143
urn:nbn:de:0074-4142-0


Vol-4142/paper16⫷Vol-4142/paper17⫸Vol-4142/paper18
Raik DankworthGesina Schwalbe

Attack logics, not outputs: Towards efficient robustification of deep neural networks by falsifying concept-based properties