Backdoor Attack with Mode Mixture Latent Modification

arXiv:2403.07463v1 Announce Type: new
Abstract: Backdoor attacks become a significant security concern for deep neural networks in recent years. An image classification model can be compromised if malicious backdoors are injected into it. This corruption will cause the model to function normally on clean images but predict a specific target label when triggers are present. Previous research can be categorized into two genres: poisoning a portion of the dataset with triggered images for users to train the model from scratch, …

arxiv attack attacks backdoor backdoor attack backdoor attacks backdoors can classification compromised corruption cs.cr cs.cv function image images malicious mode modification networks neural networks predict security target