Timezone: »

 
Spotlight
Query-Efficient and Scalable Black-Box Adversarial Attacks on Discrete Sequential Data via Bayesian Optimization
Deokjae Lee · Seungyong Moon · Junhyeok Lee · Hyun Oh Song

Wed Jul 20 10:40 AM -- 10:45 AM (PDT) @ Room 318 - 320
in DL: Robustness »

We focus on the problem of adversarial attacks against models on discrete sequential data in the black-box setting where the attacker aims to craft adversarial examples with limited query access to the victim model. Existing black-box attacks, mostly based on greedy algorithms, find adversarial examples using pre-computed key positions to perturb, which severely limits the search space and might result in suboptimal solutions. To this end, we propose a query-efficient black-box attack using Bayesian optimization, which dynamically computes important positions using an automatic relevance determination (ARD) categorical kernel. We introduce block decomposition and history subsampling techniques to improve the scalability of Bayesian optimization when an input sequence becomes long. Moreover, we develop a post-optimization algorithm that finds adversarial examples with smaller perturbation size. Experiments on natural language and protein classification tasks demonstrate that our method consistently achieves higher attack success rate with significant reduction in query count and modification rate compared to the previous state-of-the-art methods.

Keywords: APP: Language, Speech and Dialog · DL: Robustness

Author Information

Deokjae Lee (Seoul National University)
Seungyong Moon (Seoul National University)
Junhyeok Lee (Seoul National University)
Hyun Oh Song (Seoul National University)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors