Abstract: We propose a general attack framework based on evolutionary algorithms to quickly and efficiently generate low-perturbation adversarial samples for 3D point cloud data. Specifically, we ...

Abstract: Dynamic multiobjective optimization problems (DMOPs) frequently arise in real-world applications, where environments evolve over time. A central challenge in solving DMOPs is the ability to ...