Adversarial examplesDeep learningPoint cloud object detection is crucial for a variety of applications, including autonomous driving and robotics. Voxel-based representation for 3D point clouds has drawn significant attention due to their efficiency and effectiveness. Recent studies have revealed the ...
The phenomenon that deep neural networks are vulnerable to adversarial examples has been found for several years. Under the black-box setting, transfer-based methods usually produce the adversarial examples on a white-box model, which serves as the surrogate model in the black-box attack, and hop...
Results showed that all 4 detection methods provide a very robust performance while KD achieves the highest AUC of 99% against all attacks for three studied datasets. On the other hand, these detection methods obtain less than 80% AUC for real-world images. To answer why medical adversarial ...