Unauthorized AI cannot recognize me: Reversible adversarial example

作者：

Highlights：

• Reversible adversarial example controls how user’s data is recognized by AI.

• The original image can be exactly recovered from reversible adversarial example.

• Reversible adversarial example still functions as adversarial example.

摘要

•Reversible adversarial example controls how user’s data is recognized by AI.•The original image can be exactly recovered from reversible adversarial example.•Reversible adversarial example still functions as adversarial example.

论文关键词：Adversarial example,Reversible data hiding,AI security

论文评审过程：Received 28 January 2022, Revised 28 July 2022, Accepted 20 September 2022, Available online 24 September 2022, Version of Record 13 October 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.109048