Black-Box-Attacks
November 4, 2021 ยท View on GitHub
This is a black-box attack library.
Attack Methods
| Attack Methods | Attack Type | Similarity | Links |
|---|---|---|---|
| Zoo | score based | L_2 | ZOO: Zeroth Order Optimization based Black-box Attacks to Deep Neural Networks without Training Substitute Models |
| Nattack | score based | L_inf | NATTACK: Learning the Distributions of Adversarial Examples for an Improved Black-Box Attack on Deep Neural Networks |
| SimBA | score based | L_2 | Simple Black-box Adversarial Attacks |
| Bandit | score based | L_2 | Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors) |
| Boundary Attack | decision based | - | Decision-Based Adversarial Attacks: Reliable Attacks Against Black-Box Machine Learning Models |
| HSJA | decision based | - | HopSkipJumpAttack: A Query-Efficient Decision-Based Attack |
| QEBA | decision based | - | QEBA: Query-Efficient Boundary-Based Blackbox Attack |
| Bayes Attack | decison based | - | Simple Black-box Adversarial Attacks |
Datasets
Currently these attacks are only test on MNIST and CIFAR10.