Paper
in
Workshop: Security and Safety in Machine Learning Systems
PatchGuard++: Efficient Provable Attack Detection against Adversarial Patches
Chong Xiang
Abstract: