Detecting Backdoors in Object Detection via Pre-NMS Prediction Distribution Shift
DistScan is presented, a backdoor detection framework based on a simple but previously unexploited observation: backdoor injection systematically shifts a model's pre-NMS prediction class distribution away from its training class frequencies, even on clean inputs without any trigger present.
Longtian Wang, Zhengyu Zhao, Chenhao Lin et al.
· 0 citations