TnT Attacks! Universal Naturalistic Adversarial Patches Against Deep Neural Network Systems

Reference

Doan, B.G., Xue, M., Ma, S., Abbasnejad, E. and Ranasinghe, D.C., 2021. TnT Attacks! Universal Naturalistic Adversarial Patches Against Deep Neural Network Systems. arXiv preprint arXiv:2111.09999.

Link to the research paper: TnTattacks

Bibtex

@article{doan2021tnt,
  title={TnT Attacks! Universal Naturalistic Adversarial Patches Against Deep Neural Network Systems},
  author={Doan, Bao Gia and Xue, Minhui and Ma, Shiqing and Abbasnejad, Ehsan and Ranasinghe, Damith C},
  journal={arXiv preprint arXiv:2111.09999},
  year={2021}
}

Physical World Deployment Demonstration Videos

Summary

Targeted Attacks on the PubFig Classification Task

Targeted Attacks on ImageNet Large Scale Visual Recognition Task

Demo Videos

Targeted Attack with an input-agnostic TnT, an example flower, from the TnT Generator - PubFig Face Recognition Task

The effectiveness of an example flower TnT

The robustness of an example flower TnT

Targeted Attack with an Input-Agnostic flower Trigger from the TnT Generator - ImageNet classification task

The effectiveness of an example flower TnT

The effectiveness and robustness of miscellaneous examples of flower TnTs

Targeted Attack with an Input-Agnostic Patch from the Adversarial Patch Generator - ImageNet classification task

The effectiveness of a patch trigger

The robustness of a patch trigger