L-Red: Efficient Post-Training Detection of Imperceptible Backdoor Attacks Without Access to the Training Set.

Zhen Xiang, David J. Miller, George Kesidis

Browse the full ICASSP paper archive.