Data-free Defense of Black Box Models Against Adversarial Attacks

arXiv:2211.01579v3 Announce Type: replace
Abstract: Several companies often safeguard their trained deep models (i.e., details of architecture, learnt weights, training details etc.) from third-party users by exposing them only as black boxes through APIs. Moreover, they may not even provide access to the training data due to proprietary reasons or sensitivity concerns. In this work, we propose a novel defense mechanism for black box models against adversarial attacks in a data-free set up. We construct synthetic data via generative model …

adversarial adversarial attacks arxiv attacks black box box cs.cr cs.cv cs.lg data defense free type