Clean-image Backdoor Attacks

arXiv:2403.15010v1 Announce Type: new
Abstract: To gather a significant quantity of annotated training data for high-performance image classification models, numerous companies opt to enlist third-party providers to label their unlabeled data. This practice is widely regarded as secure, even in cases where some annotated errors occur, as the impact of these minor inaccuracies on the final performance of the models is negligible and existing backdoor attacks require attacker's ability to poison the training images. Nevertheless, in this paper, we propose …

abstract arxiv attacks backdoor cases classification companies cs.cr cs.cv data errors gather image impact performance practice training training data type