A data independent approach to generate adversarial patches
A data independent approach to generate adversarial patches
Xingyu Zhou,Zhisong Pan,2 Authors,Shuaihui Wang
2021 · DOI: 10.1007/s00138-021-01194-6
Machine Vision and Applications · 13 Citations
TLDR
A data independent approach to generate adversarial patches (DiAP) that can fool the target model on most of the images without any knowledge about the training data distribution, which outperforms state-of-the-art adversarial patch attack methods.