AI Paper Atlas

Research map / Adversarial robustness and security / Black-box and transfer attacks

Aha! Adaptive History-Driven Attack for Decision-Based Black-Box Models

Jie Li, Rongrong Ji, Peixian Chen, Baochang Zhang, Xiaopeng Hong, Ruixin Zhang, Shaoxin Li, Jilin Li, Feiyue Huang, Yongjian Wu

ICCV 2021 · Adversarial robustness and security › Black-box and transfer attacks · approach: Query-efficient black-box attacks · 24 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

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 changedWhereMeasuredEffect
add dimension reduction (subspace sampling) (ablation)AHA · ImageNet ResNet50mean L2: 10.823 -> 6.203Improved
add dimension reduction (subspace sampling) (ablation)AHA · ImageNet Inception-V3mean L2: 22.196 -> 10.976Improved
add DCT low-pass on top of dimension reduction (ablation)AHA · ImageNet ResNet50mean L2: 6.203 -> 16.200Hurt
use all historical queries rather than only failed ones (ablation)AHA · ImageNet ResNet50mean L2: 6.950 -> 6.203Improved
use all historical queries rather than only successful ones (ablation)AHA · ImageNet ResNet50mean L2: 7.465 -> 6.203Improved
replace 1/5th success rule with ratio-based adaptive coefficient h=(r+0.8)^2 (ablation)AHA · ImageNet ResNet50mean L2: 6.638 -> 6.203Improved
replace 1/5th success rule with ratio-based adaptive coefficient h=(r+0.8)^2 (ablation)AHA · ImageNet Inception-V3mean L2: 11.507 -> 10.976Improved
change h from (r+0.8)^2 to (r+0.6)^2 (ablation)AHA · ImageNet ResNet50mean L2: 6.203 -> 20.813Hurt
change h from (r+0.8)^2 to (r+0.9)^2 (ablation)AHA · ImageNet ResNet50mean L2: 6.203 -> 9.643Hurt
change h from (r+0.8)^2 to r+0.8 (ablation)AHA · ImageNet ResNet50mean L2: 6.203 -> 6.168No clear effect
use AHA instead of Evolutionary attackEvolutionary · ImageNet Inception-V3attack success rate: 64% -> 91%Improved
use AHA instead of Evolutionary attackEvolutionary · Face++ APImean L2: 14.544 -> 10.910Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy