SSMI: How to Make Objects of Interest Disappear without Accessing Object Detectors?

no code implementations • 22 Jun 2022 • Hui Xia, Rui Zhang, Zi Kang, Shuliang Jiang

Most black-box adversarial attack schemes for object detectors mainly face two shortcomings: requiring access to the target model and generating inefficient adversarial examples (failing to make objects disappear in large numbers).

Adversarial Attack Object +3

AdvSmo: Black-box Adversarial Attack by Smoothing Linear Structure of Texture

no code implementations • 22 Jun 2022 • Hui Xia, Rui Zhang, Shuliang Jiang, Zi Kang

We construct the adversarial examples without relying on any internal information to the target model and design the imperceptible-high attack success rate constraint to guide the Gabor filter to select appropriate angles and scales to smooth the linear texture from the input images to generate adversarial examples.

Adversarial Attack Adversarial Defense