Research map / Adversarial robustness and security / Black-box and transfer attacks
Improving Query Efficiency of Black-box Adversarial Attack
Yang Bai, Yuyuan Zeng, Yong Jiang, Yisen Wang, Shu-Tao Xia, Weiwei Guo
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace NAttack's per-pixel Gaussian with a Neural Process decoder (NP-Attack-R) | NAttack · CIFAR-10 untargeted | average query count: 183 -> 94 | Improved |
| replace NAttack with NP-Attack-R | NAttack · MNIST untargeted | average query count: 4,357 -> 1,226 | Improved |
| replace NAttack with NP-Attack-R | NAttack · MNIST targeted | average query count: 5,981 -> 2,693 | Improved |
| replace NAttack with NP-Attack-R | NAttack · ImageNet untargeted | average query count: 2,075 -> 867 | Improved |
| replace NAttack with NP-Attack-R | NAttack · ImageNet targeted | average query count: 14,229 -> 8,001 | Improved |
| replace QL with NP-Attack-R | QL · MNIST targeted | average query count: 2,693 -> 2,693 | No clear effect |
| replace NAttack with NP-Attack-R on other architectures | NAttack · CIFAR-10 ResNet18 | average query count: 311 -> 185 | Improved |
| optimize latent z too (NP-Attack-RZ) instead of r only (ablation) | NP-Attack-R · MNIST untargeted | average query count: 1,190 -> 1,460 | Hurt |
| optimize z only instead of r only (ablation) | NP-Attack-R · CIFAR-10 untargeted | average query count: 96 -> 150 | Hurt |
| increase sample size b from 10 to 50 (ablation) | NP-Attack-R · MNIST untargeted | attack success rate and query count: 98.01% -> 99.00% ASR, 723 -> 1,203 queries | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- NAttack — NATTACK: Learning the Distributions of Adversarial Examples for an Improved Black-Box Attack on Deep Neural Networks (ICML 2019) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Practical Black-box Attacks on Deep Neural Networks using Efficient Query Mechanisms (ECCV 2018 · similarity 0.92)
- Towards Query-Efficient Black-Box Adversary with Zeroth-Order Natural Gradient Descent (AAAI 2020 · similarity 0.91)
- Understanding the Robustness of Randomized Feature Defense Against Query-Based Adversarial Attacks (ICLR 2024 · similarity 0.91)
- Simple Black-box Adversarial Attacks (ICML 2019 · similarity 0.91)
- Efficient Black-box Adversarial Attacks via Bayesian Optimization Guided by a Function Prior (ICML 2024 · similarity 0.89)
- Mind the Gap: Detecting Black-box Adversarial Attacks in the Making through Query Update Analysis (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
