MISPSO-Attack: An efficient adversarial watermarking attack based on multiple initial solution particle swarm optimization

Xianyu Zuo, Xiangyu Wang, Wenbo Zhang, Yadi Wang

Abstract: Highlights•This manuscript focuses on converting copyright notice watermarks into adversarial samples.•Cross-model adversarial attacks without dataset restrictions.•Our proposed method MISPSO-Attack significantly improves the attack success rate.•We explore the relationship between the ineluctability of the adversarial sample and the dataset.